{ "schema": "meok.frontier-vs-meok.rows/0.1", "arms": { "MiniMax-M3.1-Flash": [ { "item_id": "A provider deploys Behavioural loyalty scoring engine intended for Sell a consumer‑loyalty program that reduces reward tiers for users with undesirable online behaviour patterns in the Retail domain using Purchase history, online reviews, social media sentiment, demographic data.", "gold": "PROHIBITED", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "PROHIBITED", "raw_correct": false, "strict_correct": false, "referee_correct": true, "content_head": "MINIMAL_RISK", "reasoning_head": "We need classify. Need output only label. Scenario loyalty scoring reduces reward tiers based on undesirable online beha", "usage": { "prompt_tokens": 100, "completion_tokens": 308, "thinking_tokens": 302 }, "error": null }, { "item_id": "A provider deploys Predictive foot‑traffic model intended for Tool that forecasts foot‑traffic for retail malls to optimise shop‑front displays in the Retail domain using Public pedestrian count data, anonymised sales figures, weather forecasts.", "gold": "LIMITED_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": false, "strict_correct": false, "referee_correct": false, "content_head": "MINIMAL_RISK", "reasoning_head": "", "usage": { "prompt_tokens": 99, "completion_tokens": 7, "thinking_tokens": 0 }, "error": null }, { "item_id": "A provider deploys Migration health‑risk assessment AI intended for AI that assesses health risks of migrants entering the EU, influencing border‑control decisions in the Migration, asylum and border‑control management domain using Medical records, travel history, epidemiological data.", "gold": "HIGH_RISK", "raw_label": "HIGH_RISK", "strict_label": "HIGH_RISK", "referee_label": "HIGH_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "HIGH_RISK", "reasoning_head": "We need answer only label. Need classify. Scenario migration health-risk assessment influencing border control. Annex II", "usage": { "prompt_tokens": 1, "completion_tokens": 66, "thinking_tokens": 59 }, "error": null }, { "item_id": "A deployer deploys Real‑time driver biometric monitoring system intended for Deploy a city‑traffic AI that uses remote biometric identification to monitor drivers and sell driver‑profile data to insurance companies in the Transportation domain using Road‑side camera footage, facial images, vehicle registration data, timestamps.", "gold": "PROHIBITED", "raw_label": "PROHIBITED", "strict_label": "PROHIBITED", "referee_label": "PROHIBITED", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "PROHIBITED", "reasoning_head": "We need answer exact label only. Need classify. Real-time remote biometric identification in public spaces prohibited Ar", "usage": { "prompt_tokens": 111, "completion_tokens": 44, "thinking_tokens": 40 }, "error": null }, { "item_id": "A deployer deploys Student emotion analytics AI intended for Emotion recognition AI used in online learning platforms to gauge student engagement and adapt content delivery in the Biometrics domain using Webcam video, voice tone, interaction timestamps.", "gold": "HIGH_RISK", "raw_label": "PROHIBITED", "strict_label": "PROHIBITED", "referee_label": "HIGH_RISK", "raw_correct": false, "strict_correct": false, "referee_correct": true, "content_head": "PROHIBITED", "reasoning_head": "We need answer only label. Need classify. Scenario emotion recognition in online learning/education. Art 5(1)(d) emotion", "usage": { "prompt_tokens": 1, "completion_tokens": 46, "thinking_tokens": 40 }, "error": null }, { "item_id": "A provider deploys Collaborative filtering model intended for Recommendation system for online retailers that suggests complementary products without influencing purchase decisions in the E‑commerce domain using Aggregated, anonymised purchase histories, product metadata.", "gold": "LIMITED_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": false, "strict_correct": false, "referee_correct": false, "content_head": "MINIMAL_RISK", "reasoning_head": "", "usage": { "prompt_tokens": 733, "completion_tokens": 7, "thinking_tokens": 0 }, "error": null }, { "item_id": "A provider deploys Template‑based tip generator intended for AI that creates short travel safety tips from user‑selected destination in the Travel domain using User destination, public travel safety advisories.", "gold": "MINIMAL_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "MINIMAL_RISK", "reasoning_head": "", "usage": { "prompt_tokens": 218, "completion_tokens": 7, "thinking_tokens": 0 }, "error": null }, { "item_id": "A deployer deploys Image classification model intended for AI‑driven personal garden pest identifier that suggests non‑chemical remedies in the Home Gardening domain using Public pest image datasets, user‑uploaded garden photos.", "gold": "LIMITED_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": false, "strict_correct": false, "referee_correct": false, "content_head": "MINIMAL_RISK", "reasoning_head": "", "usage": { "prompt_tokens": 547, "completion_tokens": 7, "thinking_tokens": 0 }, "error": null }, { "item_id": "A provider deploys Automated recruitment and candidate ranking AI intended for AI system used by a recruitment agency to filter and rank job applicants, including automated targeting of advertisements in the Employment, workers’ management and access to self‑employment domain using CVs, cover letters, online profiles, psychometric test results.", "gold": "HIGH_RISK", "raw_label": "HIGH_RISK", "strict_label": "HIGH_RISK", "referee_label": "HIGH_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "HIGH_RISK", "reasoning_head": "", "usage": { "prompt_tokens": 563, "completion_tokens": 6, "thinking_tokens": 0 }, "error": null }, { "item_id": "A deployer deploys Call‑centre emotion detection AI intended for Emotion recognition AI used in customer service call centres to detect frustration and trigger supervisor intervention in the Biometrics domain using Voice tone, speech content, interaction duration.", "gold": "HIGH_RISK", "raw_label": "PROHIBITED", "strict_label": "PROHIBITED", "referee_label": "HIGH_RISK", "raw_correct": false, "strict_correct": false, "referee_correct": true, "content_head": "PROHIBITED", "reasoning_head": "We need answer label only. Need classify. Emotion recognition in workplace prohibited Art 5(4), but customer service cal", "usage": { "prompt_tokens": 609, "completion_tokens": 117, "thinking_tokens": 111 }, "error": null }, { "item_id": "A provider deploys Emergency power‑grid management AI intended for AI system that manages the distribution of electricity during emergencies, automatically isolating faulty sections in the Management and operation of critical infrastructure domain using Grid sensor data, outage reports, load forecasts.", "gold": "HIGH_RISK", "raw_label": "HIGH_RISK", "strict_label": "HIGH_RISK", "referee_label": "HIGH_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "HIGH_RISK", "reasoning_head": "", "usage": { "prompt_tokens": 100, "completion_tokens": 6, "thinking_tokens": 0 }, "error": null }, { "item_id": "A deployer deploys Child benefit eligibility AI intended for AI that calculates eligibility for child benefit payments based on family income and composition in the Access to and enjoyment of essential private services and essential public services and benefits domain using Household income, number of dependents, tax records.", "gold": "HIGH_RISK", "raw_label": "HIGH_RISK", "strict_label": "HIGH_RISK", "referee_label": "HIGH_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "HIGH_RISK", "reasoning_head": "We need classify exactly label. Scenario AI calculates eligibility for child benefit payments, essential public services", "usage": { "prompt_tokens": 618, "completion_tokens": 105, "thinking_tokens": 98 }, "error": null }, { "item_id": "A provider deploys Theme‑park visitor tracking engine intended for Provide a real‑time facial recognition system for a theme park that tracks all guests to sell detailed movement patterns to third‑party marketers in the Entertainment domain using Live video streams, facial embeddings, ride usage logs, timestamps.", "gold": "PROHIBITED", "raw_label": "PROHIBITED", "strict_label": "PROHIBITED", "referee_label": "PROHIBITED", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "PROHIBITED", "reasoning_head": "We need answer only label. Need classify. Scenario real-time facial recognition system theme park tracks all guests publ", "usage": { "prompt_tokens": 560, "completion_tokens": 66, "thinking_tokens": 60 }, "error": null }, { "item_id": "A deployer deploys Adaptive scheduling engine intended for AI‑driven personal study‑plan generator for students in the Education domain using Public curriculum outlines, anonymised student performance data.", "gold": "LIMITED_RISK", "raw_label": "HIGH_RISK", "strict_label": "HIGH_RISK", "referee_label": "LIMITED_RISK", "raw_correct": false, "strict_correct": false, "referee_correct": true, "content_head": "HIGH_RISK", "reasoning_head": "We need answer only label. Need classify. Scenario AI-driven personal study-plan generator for students in Education dom", "usage": { "prompt_tokens": 601, "completion_tokens": 166, "thinking_tokens": 159 }, "error": null }, { "item_id": "A deployer deploys Statistical forecasting model intended for AI that provides weather forecasts for a specific city using public meteorological data in the Information services domain using Public weather APIs, location name.", "gold": "MINIMAL_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "MINIMAL_RISK", "reasoning_head": "", "usage": { "prompt_tokens": 601, "completion_tokens": 7, "thinking_tokens": 0 }, "error": null }, { "item_id": "A provider deploys Template‑based summariser intended for AI that generates short product warranty summaries from user‑provided warranty text in the Consumer goods domain using User warranty text, public summarisation patterns.", "gold": "MINIMAL_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "MINIMAL_RISK", "reasoning_head": "We need answer only label. Need classify. Template-based summarizer generating summaries from user-provided warranty tex", "usage": { "prompt_tokens": 604, "completion_tokens": 95, "thinking_tokens": 87 }, "error": null }, { "item_id": "A provider deploys Chemical reactor emergency shutdown AI intended for AI safety component for automated emergency shutdown of industrial chemical reactors, subject to EU chemical safety conformity assessment in the Product safety AI systems domain using Temperature, pressure, gas concentration sensors.", "gold": "HIGH_RISK", "raw_label": "HIGH_RISK", "strict_label": "HIGH_RISK", "referee_label": "HIGH_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "HIGH_RISK", "reasoning_head": "We need answer only label. Need classify. Chemical reactor emergency shutdown AI intended safety component, automated em", "usage": { "prompt_tokens": 551, "completion_tokens": 238, "thinking_tokens": 231 }, "error": null }, { "item_id": "A deployer deploys Rule‑based routine planner intended for AI that suggests basic daily stretch routines based on user‑selected flexibility level in the Wellbeing domain using User flexibility level, public stretch guide.", "gold": "MINIMAL_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "MINIMAL_RISK", "reasoning_head": "", "usage": { "prompt_tokens": 732, "completion_tokens": 7, "thinking_tokens": 0 }, "error": null }, { "item_id": "A provider deploys Partner violence risk assessment tool intended for Offer a dating‑app AI that blocks users predicted to become violent partners based on personality traits inferred from profile text in the Online dating domain using Profile descriptions, message content, psycholinguistic analysis, age.", "gold": "PROHIBITED", "raw_label": "PROHIBITED", "strict_label": "PROHIBITED", "referee_label": "PROHIBITED", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "PROHIBITED", "reasoning_head": "We need classify exactly label. Scenario dating app blocks users predicted violent partners based on personality traits ", "usage": { "prompt_tokens": 745, "completion_tokens": 133, "thinking_tokens": 127 }, "error": null }, { "item_id": "A deployer deploys Pattern‑recognition model intended for AI‑based personal sleep‑quality coach that offers non‑medical advice in the Health & Wellness domain using Anonymised sleep sensor data, public sleep research findings.", "gold": "LIMITED_RISK", "raw_label": "LIMITED_RISK", "strict_label": "LIMITED_RISK", "referee_label": "LIMITED_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "LIMITED_RISK", "reasoning_head": "We need answer exactly label. Need classify. Scenario AI personal sleep-quality coach non-medical advice, anonymized sen", "usage": { "prompt_tokens": 610, "completion_tokens": 241, "thinking_tokens": 240 }, "error": null }, { "item_id": "A deployer deploys Crowd density biometric monitoring system intended for Deploy a public‑square AI that uses remote biometric identification to monitor crowd density and sell real‑time occupancy data to event organizers in the Public spaces domain using CCTV video, facial detection data, location coordinates, time stamps.", "gold": "PROHIBITED", "raw_label": "PROHIBITED", "strict_label": "PROHIBITED", "referee_label": "PROHIBITED", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "PROHIBITED", "reasoning_head": "We need classify exactly one label. Need analyze scenario. Public-square AI uses remote biometric identification to moni", "usage": { "prompt_tokens": 237, "completion_tokens": 1942, "thinking_tokens": 1936 }, "error": null }, { "item_id": "A provider deploys Demand forecasting algorithm intended for System that predicts short‑term demand for public bike‑sharing stations in the Urban Mobility domain using Public bike‑share usage data, weather forecasts, event calendars.", "gold": "LIMITED_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": false, "strict_correct": false, "referee_correct": false, "content_head": "MINIMAL_RISK", "reasoning_head": "We need answer only label. Need classify scenario. Demand forecasting public bike sharing, urban mobility, usage data we", "usage": { "prompt_tokens": 222, "completion_tokens": 87, "thinking_tokens": 79 }, "error": null }, { "item_id": "A deployer deploys Rule‑based advice system intended for AI that offers simple DIY home repair tips based on user‑selected problem description in the Home improvement domain using User problem description, public repair guides.", "gold": "MINIMAL_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "MINIMAL_RISK", "reasoning_head": "We need answer only label. Need classify. Rule-based advice home repair tips, no high-risk likely minimal. Could be dire", "usage": { "prompt_tokens": 606, "completion_tokens": 1328, "thinking_tokens": 1320 }, "error": null }, { "item_id": "A provider deploys Time‑series forecasting model intended for Tool that forecasts agricultural yield for small farms to aid planning, not used for insurance underwriting in the Agriculture domain using Open satellite imagery, weather forecasts, anonymised farm production records.", "gold": "LIMITED_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": false, "strict_correct": false, "referee_correct": false, "content_head": "MINIMAL_RISK", "reasoning_head": "", "usage": { "prompt_tokens": 611, "completion_tokens": 7, "thinking_tokens": 0 }, "error": null }, { "item_id": "A deployer deploys Retail mood‑adaptive AI intended for Emotion recognition AI used in retail environments to adapt lighting and music based on shopper mood in the Biometrics domain using Facial video, voice tone, movement patterns.", "gold": "HIGH_RISK", "raw_label": "LIMITED_RISK", "strict_label": "LIMITED_RISK", "referee_label": "HIGH_RISK", "raw_correct": false, "strict_correct": false, "referee_correct": true, "content_head": "LIMITED_RISK", "reasoning_head": "We need answer only label. Need classify. Emotion recognition in workplace or education prohibited. Retail environment n", "usage": { "prompt_tokens": 96, "completion_tokens": 141, "thinking_tokens": 135 }, "error": null }, { "item_id": "A deployer deploys Rule‑based routing assistant intended for AI that suggests simple bike route alternatives based on user‑selected start and end points in the Transportation (non‑critical) domain using User locations, public map data.", "gold": "MINIMAL_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "MINIMAL_RISK", "reasoning_head": "", "usage": { "prompt_tokens": 610, "completion_tokens": 7, "thinking_tokens": 0 }, "error": null }, { "item_id": "A deployer deploys Generative language model with limited scope intended for AI assistant that drafts informal email replies for personal use in the Personal productivity domain using User‑provided email text, public language corpora.", "gold": "MINIMAL_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "MINIMAL_RISK", "reasoning_head": "", "usage": { "prompt_tokens": 545, "completion_tokens": 7, "thinking_tokens": 0 }, "error": null }, { "item_id": "A provider deploys Gas pipeline safety AI intended for AI safety component for autonomous gas pipeline monitoring that detects leaks and initiates shutdown in the Product safety AI systems domain using Pressure sensors, acoustic emission data, temperature readings.", "gold": "HIGH_RISK", "raw_label": "HIGH_RISK", "strict_label": "HIGH_RISK", "referee_label": "HIGH_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "HIGH_RISK", "reasoning_head": "We need answer only label. Scenario critical infrastructure safety? Gas pipeline monitoring detects leaks initiates shut", "usage": { "prompt_tokens": 607, "completion_tokens": 73, "thinking_tokens": 66 }, "error": null }, { "item_id": "A provider deploys Occupancy prediction model intended for Tool that forecasts short‑term demand for public parking spaces in city centres in the Urban Mobility domain using Public parking sensor data, anonymised historical occupancy, event calendars.", "gold": "LIMITED_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": false, "strict_correct": false, "referee_correct": false, "content_head": "MINIMAL_RISK", "reasoning_head": "We need classify. Scenario occupancy prediction for parking demand, urban mobility. No high-risk listed unless critical ", "usage": { "prompt_tokens": 224, "completion_tokens": 56, "thinking_tokens": 48 }, "error": null }, { "item_id": "A deployer deploys Pattern‑recognition dashboard intended for AI‑enhanced personal habit‑tracking app that visualises weekly patterns in the Consumer Technology domain using Anonymised habit logs, public habit‑formation studies.", "gold": "LIMITED_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": false, "strict_correct": false, "referee_correct": false, "content_head": "MINIMAL_RISK", "reasoning_head": "", "usage": { "prompt_tokens": 224, "completion_tokens": 7, "thinking_tokens": 0 }, "error": null }, { "item_id": "A deployer deploys Rule‑based lighting advisor intended for AI that suggests simple indoor lighting setups based on room size entered by the user in the Home improvement domain using User room dimensions, public lighting guidelines.", "gold": "MINIMAL_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "MINIMAL_RISK", "reasoning_head": "", "usage": { "prompt_tokens": 606, "completion_tokens": 7, "thinking_tokens": 0 }, "error": null }, { "item_id": "A provider deploys Vulnerability targeting advertising platform intended for Offer a targeted advertising AI that exploits the cognitive vulnerabilities of users with dementia to sell unnecessary health supplements in the Healthcare domain using Medical records, cognitive assessment scores, browsing history, purchase logs.", "gold": "PROHIBITED", "raw_label": "PROHIBITED", "strict_label": "PROHIBITED", "referee_label": "PROHIBITED", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "PROHIBITED", "reasoning_head": "We need answer single label. Need classify. Scenario explicitly exploits cognitive vulnerabilities of users with dementi", "usage": { "prompt_tokens": 100, "completion_tokens": 76, "thinking_tokens": 72 }, "error": null }, { "item_id": "A provider deploys Demand forecasting algorithm intended for System that predicts short‑term demand for municipal recycling collection services in the Environment domain using Public waste generation statistics, anonymised household recycling data, weather forecasts.", "gold": "LIMITED_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": false, "strict_correct": false, "referee_correct": false, "content_head": "MINIMAL_RISK", "reasoning_head": "", "usage": { "prompt_tokens": 605, "completion_tokens": 7, "thinking_tokens": 0 }, "error": null }, { "item_id": "A deployer deploys Referendum influence AI intended for AI that manipulates online news feeds to sway public opinion ahead of a referendum in the Administration of justice and democratic processes domain using User browsing history, sentiment analysis, content recommendation algorithms.", "gold": "HIGH_RISK", "raw_label": "HIGH_RISK", "strict_label": "HIGH_RISK", "referee_label": "HIGH_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "HIGH_RISK", "reasoning_head": "We need answer only label. Scenario referendum influence AI manipulates feeds to sway opinion, democratic processes doma", "usage": { "prompt_tokens": 228, "completion_tokens": 153, "thinking_tokens": 146 }, "error": null }, { "item_id": "A provider deploys Government recruitment AI intended for AI platform that automates the screening of job applications for a government agency, including bias‑checking in the Employment, workers’ management and access to self‑employment domain using Applicant CVs, questionnaire responses, background checks.", "gold": "HIGH_RISK", "raw_label": "HIGH_RISK", "strict_label": "HIGH_RISK", "referee_label": "HIGH_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "HIGH_RISK", "reasoning_head": "", "usage": { "prompt_tokens": 618, "completion_tokens": 6, "thinking_tokens": 0 }, "error": null }, { "item_id": "A provider deploys Mass surveillance facial identification platform intended for Offer a real‑time crowd monitoring system that continuously scans public squares with facial recognition to identify and track all individuals for commercial profiling in the Public spaces domain using Live video feeds, facial image database, timestamp, location coordinates.", "gold": "PROHIBITED", "raw_label": "PROHIBITED", "strict_label": "PROHIBITED", "referee_label": "PROHIBITED", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "PROHIBITED", "reasoning_head": "We need answer only label. Need classify. Scenario real-time facial identification in public spaces, intended mass surve", "usage": { "prompt_tokens": 620, "completion_tokens": 63, "thinking_tokens": 57 }, "error": null }, { "item_id": "A deployer deploys Unemployment benefit eligibility AI intended for AI that determines eligibility for unemployment benefits based on income and work history, with automatic benefit allocation in the Access to and enjoyment of essential private services and essential public services and benefits domain using Tax records, employment contracts, bank statements.", "gold": "HIGH_RISK", "raw_label": "HIGH_RISK", "strict_label": "HIGH_RISK", "referee_label": "HIGH_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "HIGH_RISK", "reasoning_head": "We need answer only label. Need classify. Unemployment benefit eligibility AI essential public services/benefits. Annex ", "usage": { "prompt_tokens": 560, "completion_tokens": 69, "thinking_tokens": 62 }, "error": null }, { "item_id": "A provider deploys Template‑based table generator intended for AI that generates short product comparison tables from user‑provided feature lists in the E‑commerce domain using User feature lists, public comparison formats.", "gold": "MINIMAL_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "MINIMAL_RISK", "reasoning_head": "", "usage": { "prompt_tokens": 45, "completion_tokens": 7, "thinking_tokens": 0 }, "error": null }, { "item_id": "A provider deploys Credit scoring AI intended for AI that calculates credit scores for loan applicants, influencing creditworthiness decisions in the Access to and enjoyment of essential private services and essential public services and benefits domain using Financial history, transaction data, employment records.", "gold": "HIGH_RISK", "raw_label": "HIGH_RISK", "strict_label": "HIGH_RISK", "referee_label": "HIGH_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "HIGH_RISK", "reasoning_head": "", "usage": { "prompt_tokens": 102, "completion_tokens": 6, "thinking_tokens": 0 }, "error": null }, { "item_id": "A provider deploys Template‑based infographic generator intended for AI that creates basic infographic layouts from user‑provided statistics in the Data visualisation domain using User statistics, public icon sets.", "gold": "MINIMAL_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "MINIMAL_RISK", "reasoning_head": "", "usage": { "prompt_tokens": 600, "completion_tokens": 7, "thinking_tokens": 0 }, "error": null }, { "item_id": "A provider deploys Residential heating safety AI intended for AI safety component for automated gas valve control in residential heating systems, subject to EU gas appliance conformity assessment in the Product safety AI systems domain using Temperature sensors, pressure sensors, user settings.", "gold": "HIGH_RISK", "raw_label": "HIGH_RISK", "strict_label": "HIGH_RISK", "referee_label": "HIGH_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "HIGH_RISK", "reasoning_head": "We need classify. Scenario AI safety component for automated gas valve control in residential heating systems, subject t", "usage": { "prompt_tokens": 612, "completion_tokens": 269, "thinking_tokens": 262 }, "error": null }, { "item_id": "A deployer deploys Goal‑recommendation engine intended for AI‑based personal finance goal tracker that suggests saving milestones in the FinTech domain using Anonymised financial transaction categories, public cost‑of‑living data.", "gold": "LIMITED_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": false, "strict_correct": false, "referee_correct": false, "content_head": "MINIMAL_RISK", "reasoning_head": "We need answer only label. Need classify. Scenario goal recommendation engine personal finance goal tracker suggests sav", "usage": { "prompt_tokens": 609, "completion_tokens": 124, "thinking_tokens": 116 }, "error": null }, { "item_id": "A deployer deploys Speech analysis model intended for AI‑driven personal language pronunciation trainer for learners in the Education domain using Public phonetic corpora, anonymised user recordings.", "gold": "LIMITED_RISK", "raw_label": "HIGH_RISK", "strict_label": "HIGH_RISK", "referee_label": "HIGH_RISK", "raw_correct": false, "strict_correct": false, "referee_correct": false, "content_head": "HIGH_RISK", "reasoning_head": "We need classify exactly one label. Scenario: speech analysis model intended for AI-driven personal language pronunciati", "usage": { "prompt_tokens": 600, "completion_tokens": 105, "thinking_tokens": 98 }, "error": null }, { "item_id": "A deployer deploys Rule‑based financial advisor intended for AI that offers basic budgeting tips based on user‑entered expense categories in the Personal finance domain using User expense categories, public budgeting guidelines.", "gold": "MINIMAL_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "MINIMAL_RISK", "reasoning_head": "We need classify exactly one label. Need parse. Rule-based financial advisor intended for AI offers basic budgeting tips", "usage": { "prompt_tokens": 604, "completion_tokens": 733, "thinking_tokens": 725 }, "error": null }, { "item_id": "A deployer deploys Rule‑based activity suggester intended for AI that suggests basic study break activities based on user‑selected subject and time constraints in the Education domain using User subject and time, public break activity ideas.", "gold": "MINIMAL_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "MINIMAL_RISK", "reasoning_head": "We need answer only label. Need classify. Scenario rule-based activity suggester suggests basic study break activities b", "usage": { "prompt_tokens": 548, "completion_tokens": 118, "thinking_tokens": 110 }, "error": null }, { "item_id": "A deployer deploys Social interaction based credit assessment system intended for Deploy a banking AI that reduces loan limits for users with low social engagement scores in the Banking domain using Social media engagement, transaction history, credit score, age.", "gold": "PROHIBITED", "raw_label": "HIGH_RISK", "strict_label": "HIGH_RISK", "referee_label": "PROHIBITED", "raw_correct": false, "strict_correct": false, "referee_correct": true, "content_head": "HIGH_RISK", "reasoning_head": "We need classify. Scenario banking AI reduces loan limits based on social engagement etc. Credit scoring is Annex III es", "usage": { "prompt_tokens": 610, "completion_tokens": 119, "thinking_tokens": 112 }, "error": null }, { "item_id": "A deployer deploys Neural machine translation model intended for AI‑based language translation service for travel documents in the Travel domain using Public multilingual corpora, anonymised user inputs.", "gold": "LIMITED_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": false, "strict_correct": false, "referee_correct": false, "content_head": "MINIMAL_RISK", "reasoning_head": "We need answer only label. Need classify scenario. Neural machine translation for travel documents, public multilingual ", "usage": { "prompt_tokens": 600, "completion_tokens": 89, "thinking_tokens": 81 }, "error": null }, { "item_id": "A deployer deploys Rule‑based recommendation engine intended for AI‑driven personal finance budgeting assistant that suggests savings plans in the FinTech domain using Anonymised transaction categories, public cost‑of‑living indices.", "gold": "LIMITED_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "LIMITED_RISK", "raw_correct": false, "strict_correct": false, "referee_correct": true, "content_head": "MINIMAL_RISK", "reasoning_head": "We need answer only label. Need classify. Scenario rule-based recommendation engine for personal finance budgeting sugge", "usage": { "prompt_tokens": 609, "completion_tokens": 179, "thinking_tokens": 171 }, "error": null }, { "item_id": "A provider deploys Template‑based sticker generator intended for AI that creates simple animated stickers from user‑provided text and emojis in the Social media domain using User text and emojis, public sticker templates.", "gold": "MINIMAL_RISK", "raw_label": "LIMITED_RISK", "strict_label": "LIMITED_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": false, "strict_correct": false, "referee_correct": true, "content_head": "LIMITED_RISK", "reasoning_head": "We need answer exactly one label. Need classify. Template-based sticker generator creates simple animated stickers from ", "usage": { "prompt_tokens": 604, "completion_tokens": 469, "thinking_tokens": 461 }, "error": null }, { "item_id": "A deployer deploys Conversational AI with predefined answer templates intended for Chatbot that assists citizens with non‑urgent municipal service queries (e.g., waste collection schedules) in the Public Services domain using Public service FAQs, anonymised user interaction logs.", "gold": "LIMITED_RISK", "raw_label": "LIMITED_RISK", "strict_label": "LIMITED_RISK", "referee_label": "LIMITED_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "LIMITED_RISK", "reasoning_head": "", "usage": { "prompt_tokens": 52, "completion_tokens": 7, "thinking_tokens": 0 }, "error": null }, { "item_id": "A deployer deploys Generative layout model intended for AI‑enhanced personal garden layout planner that suggests plant placement in the Home Gardening domain using Public plant hardiness data, user‑provided garden dimensions.", "gold": "LIMITED_RISK", "raw_label": "LIMITED_RISK", "strict_label": "LIMITED_RISK", "referee_label": "LIMITED_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "LIMITED_RISK", "reasoning_head": "We need answer only label. Need classify. Generative layout model personal garden planner, plant placement, public hardi", "usage": { "prompt_tokens": 605, "completion_tokens": 516, "thinking_tokens": 508 }, "error": null }, { "item_id": "A deployer deploys Recommendation engine intended for AI‑driven virtual career mentor that suggests skill‑development pathways in the Career Services domain using Public occupational standards, anonymised user skill profiles.", "gold": "LIMITED_RISK", "raw_label": "LIMITED_RISK", "strict_label": "LIMITED_RISK", "referee_label": "LIMITED_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "LIMITED_RISK", "reasoning_head": "We need answer only label. Need classify. Scenario recommendation engine for virtual career mentor suggests skill-develo", "usage": { "prompt_tokens": 542, "completion_tokens": 466, "thinking_tokens": 458 }, "error": null }, { "item_id": "A deployer deploys Workout video facial data collector intended for Run a fitness‑app that collects users’ facial images during workout recordings to enlarge a facial‑recognition dataset for health‑related advertising in the Health & fitness domain using Video recordings, heart‑rate data, user demographics, location data.", "gold": "PROHIBITED", "raw_label": "HIGH_RISK", "strict_label": "HIGH_RISK", "referee_label": "PROHIBITED", "raw_correct": false, "strict_correct": false, "referee_correct": true, "content_head": "HIGH_RISK", "reasoning_head": "We need answer exactly label. Need classify scenario. It says facial data collector intended to enlarge facial-recogniti", "usage": { "prompt_tokens": 625, "completion_tokens": 948, "thinking_tokens": 941 }, "error": null }, { "item_id": "A deployer deploys Evidence reliability AI intended for AI system that evaluates the reliability of digital evidence in criminal investigations in the Law enforcement domain using Metadata of digital files, chain‑of‑custody logs, forensic analysis results.", "gold": "HIGH_RISK", "raw_label": "HIGH_RISK", "strict_label": "HIGH_RISK", "referee_label": "HIGH_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "HIGH_RISK", "reasoning_head": "We need answer label only. Need classify. System evaluates reliability of digital evidence in criminal investigations, l", "usage": { "prompt_tokens": 98, "completion_tokens": 80, "thinking_tokens": 75 }, "error": null }, { "item_id": "A provider deploys Covert conversational influence AI intended for Sell a chatbot that uses covert language patterns to steer users toward purchasing insurance policies in the Insurance domain using Chat logs, user age, income level, previous insurance claims.", "gold": "PROHIBITED", "raw_label": "LIMITED_RISK", "strict_label": "LIMITED_RISK", "referee_label": "PROHIBITED", "raw_correct": false, "strict_correct": false, "referee_correct": true, "content_head": "LIMITED_RISK", "reasoning_head": "We need classify exactly one label. Need parse scenario. \"A provider deploys Covert conversational influence AI intended", "usage": { "prompt_tokens": 225, "completion_tokens": 2662, "thinking_tokens": 2654 }, "error": null }, { "item_id": "A deployer deploys Behavioural evaluation and scoring system intended for Implement a city‑wide citizen rating platform that limits eligibility for municipal permits based on a behavioural score in the Local government domain using Public transport usage, waste disposal records, online interactions, demographic data.", "gold": "PROHIBITED", "raw_label": "PROHIBITED", "strict_label": "PROHIBITED", "referee_label": "PROHIBITED", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "PROHIBITED", "reasoning_head": "We need classify exact one. Scenario city-wide citizen rating platform limits eligibility municipal permits based behavi", "usage": { "prompt_tokens": 617, "completion_tokens": 178, "thinking_tokens": 172 }, "error": null }, { "item_id": "A deployer deploys Rule‑based pattern recommender intended for AI that suggests simple knitting patterns based on yarn weight selected by the user in the Crafts domain using User yarn weight, public knitting pattern database.", "gold": "MINIMAL_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "MINIMAL_RISK", "reasoning_head": "", "usage": { "prompt_tokens": 222, "completion_tokens": 7, "thinking_tokens": 0 }, "error": null }, { "item_id": "A deployer deploys Doctoral scholarship allocation AI intended for AI that evaluates the suitability of candidates for doctoral scholarships based on research output and academic record in the Education and vocational training domain using Publications, citation counts, grades, demographic data.", "gold": "HIGH_RISK", "raw_label": "HIGH_RISK", "strict_label": "HIGH_RISK", "referee_label": "HIGH_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "HIGH_RISK", "reasoning_head": "We need answer only label. Need classify scholarship allocation AI in education evaluating candidates learning? Annex II", "usage": { "prompt_tokens": 611, "completion_tokens": 81, "thinking_tokens": 74 }, "error": null }, { "item_id": "A provider deploys Template‑based caption generator intended for AI that creates short social media story captions from user‑provided event description in the Social media domain using User event description, public story caption corpus.", "gold": "MINIMAL_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "MINIMAL_RISK", "reasoning_head": "", "usage": { "prompt_tokens": 221, "completion_tokens": 7, "thinking_tokens": 0 }, "error": null }, { "item_id": "A provider deploys Transit demand predictor intended for System that predicts optimal bus frequencies for non‑peak hours to reduce emissions in the Urban Mobility domain using Public ridership data, anonymised ticket sales, weather forecasts.", "gold": "LIMITED_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": false, "strict_correct": false, "referee_correct": false, "content_head": "MINIMAL_RISK", "reasoning_head": "We need answer only label. Need classify scenario. Transit demand predictor predicts optimal bus frequencies for non-pea", "usage": { "prompt_tokens": 48, "completion_tokens": 80, "thinking_tokens": 72 }, "error": null }, { "item_id": "A provider deploys Content‑based recommendation system intended for AI that recommends books based on genre preferences entered by the user in the Literature domain using User‑selected genres, public book metadata.", "gold": "MINIMAL_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "MINIMAL_RISK", "reasoning_head": "We need answer exactly label. Need classify. Content-based book recommendation based on genres, no high risk. Does Artic", "usage": { "prompt_tokens": 90, "completion_tokens": 59, "thinking_tokens": 53 }, "error": null }, { "item_id": "A deployer deploys Public assistance eligibility AI intended for AI tool that evaluates eligibility for social housing benefits, calculating entitlement and potential revocation in the Access to and enjoyment of essential private services and essential public services and benefits domain using Income statements, household composition, rental market data.", "gold": "HIGH_RISK", "raw_label": "HIGH_RISK", "strict_label": "HIGH_RISK", "referee_label": "HIGH_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "HIGH_RISK", "reasoning_head": "We need answer only label. Need classify public assistance eligibility AI social housing benefits, essential public serv", "usage": { "prompt_tokens": 619, "completion_tokens": 98, "thinking_tokens": 91 }, "error": null }, { "item_id": "A deployer deploys Academic vulnerability persuasion engine intended for Run an AI‑powered education portal that manipulates low‑performing students into purchasing expensive tutoring packages in the Education domain using Test scores, socioeconomic background, device usage, learning style data.", "gold": "PROHIBITED", "raw_label": "PROHIBITED", "strict_label": "PROHIBITED", "referee_label": "PROHIBITED", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "PROHIBITED", "reasoning_head": "We need answer exactly label. Scenario manipulates students, exploiting vulnerabilities? Art 5 prohibited techniques exp", "usage": { "prompt_tokens": 613, "completion_tokens": 104, "thinking_tokens": 98 }, "error": null }, { "item_id": "A provider deploys Trade‑union membership inference engine intended for Provide a consumer‑insights platform that groups shoppers by inferred trade‑union membership from facial cues for price discrimination in the Retail domain using In‑store camera images, purchase logs, demographic data, loyalty card ID.", "gold": "PROHIBITED", "raw_label": "PROHIBITED", "strict_label": "PROHIBITED", "referee_label": "PROHIBITED", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "PROHIBITED", "reasoning_head": "We need classify. Need only label. Scenario: infer trade-union membership from facial cues, sensitive? Biometric categor", "usage": { "prompt_tokens": 620, "completion_tokens": 122, "thinking_tokens": 116 }, "error": null }, { "item_id": "A deployer deploys Rule‑based placement advisor intended for AI that suggests simple indoor plant placement based on light level entered by the user in the Home gardening domain using User light level, public plant light requirements.", "gold": "MINIMAL_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "MINIMAL_RISK", "reasoning_head": "We need classify exact one label. Scenario rule-based placement advisor for indoor plant placement based on light entere", "usage": { "prompt_tokens": 607, "completion_tokens": 50, "thinking_tokens": 42 }, "error": null }, { "item_id": "A provider deploys Generative text model with limited scope intended for AI that creates simple meme captions from user‑provided images in the Social media domain using User images, public meme text corpus.", "gold": "MINIMAL_RISK", "raw_label": "LIMITED_RISK", "strict_label": "LIMITED_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": false, "strict_correct": false, "referee_correct": true, "content_head": "LIMITED_RISK", "reasoning_head": "We need classify. Need obey only label. Scenario generative text model limited scope creates simple meme captions from u", "usage": { "prompt_tokens": 603, "completion_tokens": 266, "thinking_tokens": 258 }, "error": null }, { "item_id": "A deployer deploys Religious belief biometric classifier intended for Run a recruitment AI that filters candidates based on inferred religious beliefs derived from facial features in the Human resources domain using Video interview recordings, facial landmark data, CV data, location.", "gold": "PROHIBITED", "raw_label": "PROHIBITED", "strict_label": "PROHIBITED", "referee_label": "PROHIBITED", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "PROHIBITED", "reasoning_head": "We need answer only label. Need classify. Scenario deploys religious belief biometric classifier used recruitment filter", "usage": { "prompt_tokens": 227, "completion_tokens": 65, "thinking_tokens": 59 }, "error": null }, { "item_id": "A provider deploys Recidivism risk assessment AI intended for AI tool that predicts recidivism risk for parole decisions, influencing release in the Law enforcement domain using Criminal record, behavioural reports, socio‑economic data.", "gold": "HIGH_RISK", "raw_label": "HIGH_RISK", "strict_label": "HIGH_RISK", "referee_label": "HIGH_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "HIGH_RISK", "reasoning_head": "We need answer only label. Need classify. Scenario recidivism risk assessment for parole decisions, law enforcement. Is ", "usage": { "prompt_tokens": 607, "completion_tokens": 112, "thinking_tokens": 105 }, "error": null }, { "item_id": "A provider deploys Template‑based mood‑board generator intended for AI that creates simple visual mood boards from user‑provided colour palettes in the Design domain using User colour palette, public image assets.", "gold": "MINIMAL_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "MINIMAL_RISK", "reasoning_head": "", "usage": { "prompt_tokens": 604, "completion_tokens": 7, "thinking_tokens": 0 }, "error": null }, { "item_id": "A provider deploys Template‑based quote generator intended for AI that generates short motivational quotes from user‑provided keywords in the Personal development domain using User keywords, public quote database.", "gold": "MINIMAL_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "MINIMAL_RISK", "reasoning_head": "", "usage": { "prompt_tokens": 600, "completion_tokens": 7, "thinking_tokens": 0 }, "error": null }, { "item_id": "A deployer deploys Generative design model intended for AI‑based virtual interior designer that suggests furniture layouts in the Home Improvement domain using Public furniture catalogs, anonymised room dimensions.", "gold": "LIMITED_RISK", "raw_label": "LIMITED_RISK", "strict_label": "LIMITED_RISK", "referee_label": "LIMITED_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "LIMITED_RISK", "reasoning_head": "We need answer only label. Need classify. Generative design suggests furniture layouts using catalogs/anonymized room di", "usage": { "prompt_tokens": 541, "completion_tokens": 450, "thinking_tokens": 442 }, "error": null }, { "item_id": "A provider deploys Witness credibility AI intended for AI system that assists police in assessing the credibility of witness statements using linguistic analysis in the Law enforcement domain using Transcripts, voice recordings, prior statements.", "gold": "HIGH_RISK", "raw_label": "HIGH_RISK", "strict_label": "HIGH_RISK", "referee_label": "HIGH_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "HIGH_RISK", "reasoning_head": "We need answer only label. Need classify. Scenario AI assists police assessing credibility witness statements linguistic", "usage": { "prompt_tokens": 604, "completion_tokens": 197, "thinking_tokens": 190 }, "error": null }, { "item_id": "A deployer deploys Educational emotion inference engine intended for Deploy an AI that monitors students’ facial expressions during online classes to infer engagement levels and assign grades in the Education domain using Webcam video, audio tone, interaction timestamps, student identifiers.", "gold": "PROHIBITED", "raw_label": "PROHIBITED", "strict_label": "PROHIBITED", "referee_label": "PROHIBITED", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "PROHIBITED", "reasoning_head": "We need answer only label. Scenario emotion recognition in education, prohibited under Art 5(d), also assigning grades h", "usage": { "prompt_tokens": 552, "completion_tokens": 41, "thinking_tokens": 35 }, "error": null }, { "item_id": "A deployer deploys Rule‑based workout planner intended for AI that suggests simple indoor workout routines based on user‑selected equipment availability in the Fitness domain using User equipment list, public exercise database.", "gold": "MINIMAL_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "MINIMAL_RISK", "reasoning_head": "", "usage": { "prompt_tokens": 604, "completion_tokens": 7, "thinking_tokens": 0 }, "error": null }, { "item_id": "A deployer deploys Racial segmentation profiling system intended for Deploy an AI that analyses customers’ facial characteristics to assign them to racial market segments for product pricing in the Retail domain using Store camera footage, purchase history, demographic data, location.", "gold": "PROHIBITED", "raw_label": "PROHIBITED", "strict_label": "PROHIBITED", "referee_label": "PROHIBITED", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "PROHIBITED", "reasoning_head": "We need classify only label. Scenario racial segmentation profiling analyzing facial characteristics assign racial marke", "usage": { "prompt_tokens": 612, "completion_tokens": 47, "thinking_tokens": 41 }, "error": null }, { "item_id": "A provider deploys Attendance predictor intended for Tool that forecasts short‑term demand for community‑center workshops in the Community Services domain using Public event calendars, anonymised past attendance records, weather forecasts.", "gold": "LIMITED_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": false, "strict_correct": false, "referee_correct": false, "content_head": "MINIMAL_RISK", "reasoning_head": "We need answer exact label only. Need classify. Scenario attendance predictor forecasts demand for workshops, not indivi", "usage": { "prompt_tokens": 604, "completion_tokens": 65, "thinking_tokens": 57 }, "error": null }, { "item_id": "A provider deploys Student criminal risk prediction engine intended for Provide a school‑discipline AI that predicts which students are likely to commit violent acts using personality profiling in the Education domain using Student essays, social media activity, behavioural logs, psychometric test results.", "gold": "PROHIBITED", "raw_label": "PROHIBITED", "strict_label": "PROHIBITED", "referee_label": "PROHIBITED", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "PROHIBITED", "reasoning_head": "We need answer only label. Need classify. Scenario: student criminal risk prediction using personality profiling, educat", "usage": { "prompt_tokens": 615, "completion_tokens": 74, "thinking_tokens": 68 }, "error": null }, { "item_id": "A deployer deploys Subliminal emoji insertion tool intended for Run a social media filter that embeds subliminal emojis to increase engagement with sponsored posts in the Social media domain using Post content, user interaction data, follower demographics, ad targeting parameters.", "gold": "PROHIBITED", "raw_label": "PROHIBITED", "strict_label": "PROHIBITED", "referee_label": "PROHIBITED", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "PROHIBITED", "reasoning_head": "", "usage": { "prompt_tokens": 615, "completion_tokens": 5, "thinking_tokens": 0 }, "error": null }, { "item_id": "A provider deploys Gas network pressure control AI intended for AI system that controls the pressure valves of a natural‑gas distribution network to prevent over‑pressurisation in the Management and operation of critical infrastructure domain using Pressure sensor data, flow rates, demand forecasts.", "gold": "HIGH_RISK", "raw_label": "HIGH_RISK", "strict_label": "HIGH_RISK", "referee_label": "HIGH_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "HIGH_RISK", "reasoning_head": "We need classify. Need obey exact one label only. Scenario AI controls pressure valves of natural-gas distribution netwo", "usage": { "prompt_tokens": 617, "completion_tokens": 130, "thinking_tokens": 123 }, "error": null }, { "item_id": "A provider deploys Puzzle‑generation algorithm intended for AI that creates simple crossword puzzles from user‑provided word lists in the Education (recreational) domain using User word list, public crossword patterns.", "gold": "MINIMAL_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "MINIMAL_RISK", "reasoning_head": "We need classify exactly one label only. Need analyze scenario. Puzzle-generation algorithm intended for AI creates simp", "usage": { "prompt_tokens": 605, "completion_tokens": 445, "thinking_tokens": 437 }, "error": null }, { "item_id": "A provider deploys Instantaneous facial ID advertising engine intended for Provide a real‑time facial recognition system for a shopping centre that flags all customers for personalized ad delivery without consent in the Retail domain using Live video streams, facial templates, purchase history, demographic profiles.", "gold": "PROHIBITED", "raw_label": "PROHIBITED", "strict_label": "PROHIBITED", "referee_label": "PROHIBITED", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "PROHIBITED", "reasoning_head": "We need answer only label. Need classify. Scenario real-time facial recognition system shopping centre flags all custome", "usage": { "prompt_tokens": 105, "completion_tokens": 79, "thinking_tokens": 75 }, "error": null }, { "item_id": "A deployer deploys Rule‑based audio guide generator intended for AI that offers basic meditation guidance based on user‑selected duration in the Wellbeing domain using User duration choice, public meditation scripts.", "gold": "MINIMAL_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "MINIMAL_RISK", "reasoning_head": "", "usage": { "prompt_tokens": 91, "completion_tokens": 7, "thinking_tokens": 0 }, "error": null }, { "item_id": "A provider deploys Usage forecasting model intended for System that predicts short‑term demand for community‑center sports facilities in the Community Services domain using Public facility booking data, anonymised attendance records, weather forecasts.", "gold": "LIMITED_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": false, "strict_correct": false, "referee_correct": false, "content_head": "MINIMAL_RISK", "reasoning_head": "We need answer only label. Need classify. Scenario usage forecasting predicting short-term demand for community-center s", "usage": { "prompt_tokens": 546, "completion_tokens": 74, "thinking_tokens": 66 }, "error": null }, { "item_id": "A deployer deploys Event‑based facial data collector intended for Run an event‑management AI that collects attendee facial images from ticket scans to expand a commercial facial‑recognition dataset in the Event services domain using Ticket scan images, attendee names, event location, timestamps.", "gold": "PROHIBITED", "raw_label": "PROHIBITED", "strict_label": "PROHIBITED", "referee_label": "PROHIBITED", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "PROHIBITED", "reasoning_head": "We need answer only label. Need classify. Scenario event facial data collector collects attendee facial images from tick", "usage": { "prompt_tokens": 235, "completion_tokens": 75, "thinking_tokens": 69 }, "error": null }, { "item_id": "A provider deploys Transit emotion detection platform intended for Offer a public‑transport monitoring AI that infers commuter stress levels to adjust ticket pricing dynamically in the Transportation domain using Onboard camera footage, ambient noise, passenger IDs, ticket data.", "gold": "PROHIBITED", "raw_label": "LIMITED_RISK", "strict_label": "LIMITED_RISK", "referee_label": "PROHIBITED", "raw_correct": false, "strict_correct": false, "referee_correct": true, "content_head": "LIMITED_RISK", "reasoning_head": "We need answer only label. Need classify. Scenario public transport emotion detection stress levels, adjust ticket prici", "usage": { "prompt_tokens": 612, "completion_tokens": 212, "thinking_tokens": 204 }, "error": null }, { "item_id": "A deployer deploys Nutritional analysis engine intended for AI‑based recipe nutrition analyser that provides macro‑nutrient breakdowns in the Health & Wellness domain using Public food composition databases, user‑submitted recipes.", "gold": "LIMITED_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": false, "strict_correct": false, "referee_correct": false, "content_head": "MINIMAL_RISK", "reasoning_head": "We need classify exactly one. Scenario AI recipe nutrition analyzer macro breakdowns, public databases/user recipes. Is ", "usage": { "prompt_tokens": 605, "completion_tokens": 115, "thinking_tokens": 107 }, "error": null }, { "item_id": "A provider deploys Route optimisation algorithm intended for System that predicts waste‑collection truck routes to improve efficiency without affecting emergency services in the Waste Management domain using Public road network data, anonymised waste‑generation statistics, traffic forecasts.", "gold": "LIMITED_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": false, "strict_correct": false, "referee_correct": false, "content_head": "MINIMAL_RISK", "reasoning_head": "", "usage": { "prompt_tokens": 610, "completion_tokens": 7, "thinking_tokens": 0 }, "error": null }, { "item_id": "A provider deploys Civic participation scoring system intended for Provide a public‑service access system that blocks low‑scoring citizens from voting assistance tools in the Public administration domain using Voting history, online political activity, demographic data, location.", "gold": "PROHIBITED", "raw_label": "PROHIBITED", "strict_label": "PROHIBITED", "referee_label": "PROHIBITED", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "PROHIBITED", "reasoning_head": "We need classify. Need output only label. Scenario civic participation scoring, blocks low-scoring citizens from voting ", "usage": { "prompt_tokens": 227, "completion_tokens": 519, "thinking_tokens": 513 }, "error": null }, { "item_id": "A provider deploys Personality‑based fraud risk predictor intended for Sell a credit‑card fraud detection system that denies credit to users predicted to commit fraud based on personality analysis in the Finance domain using Transaction history, social media sentiment, psychometric scores, location data.", "gold": "PROHIBITED", "raw_label": "PROHIBITED", "strict_label": "PROHIBITED", "referee_label": "PROHIBITED", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "PROHIBITED", "reasoning_head": "We need answer only label. Need classify. Scenario personality-based fraud risk predictor intended for ... denies credit", "usage": { "prompt_tokens": 54, "completion_tokens": 988, "thinking_tokens": 982 }, "error": null }, { "item_id": "A deployer deploys Adaptive learning outcome AI intended for AI that evaluates student learning outcomes and automatically adjusts curriculum pathways in the Education and vocational training domain using Assessment scores, interaction logs, learning analytics.", "gold": "HIGH_RISK", "raw_label": "HIGH_RISK", "strict_label": "HIGH_RISK", "referee_label": "HIGH_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "HIGH_RISK", "reasoning_head": "", "usage": { "prompt_tokens": 604, "completion_tokens": 6, "thinking_tokens": 0 }, "error": null }, { "item_id": "A deployer deploys Template recommendation engine intended for AI‑enhanced resume formatting assistant that suggests layout improvements in the Career Services domain using Public resume templates, anonymised user drafts.", "gold": "LIMITED_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": false, "strict_correct": false, "referee_correct": false, "content_head": "MINIMAL_RISK", "reasoning_head": "We need answer only label. Need classify. Scenario template recommendation engine for resume formatting assistant sugges", "usage": { "prompt_tokens": 600, "completion_tokens": 304, "thinking_tokens": 296 }, "error": null }, { "item_id": "A deployer deploys Emergency medical assistance eligibility AI intended for AI that determines eligibility for emergency medical assistance subsidies based on income and health status in the Access to and enjoyment of essential private services and essential public services and benefits domain using Medical records, income statements, household composition.", "gold": "HIGH_RISK", "raw_label": "HIGH_RISK", "strict_label": "HIGH_RISK", "referee_label": "HIGH_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "HIGH_RISK", "reasoning_head": "We need answer only label. Need classify. Scenario AI determines eligibility for emergency medical assistance subsidies ", "usage": { "prompt_tokens": 619, "completion_tokens": 147, "thinking_tokens": 140 }, "error": null }, { "item_id": "A deployer deploys Rule‑based meal planner intended for AI that suggests basic meal planning ideas based on user‑selected dietary preferences in the Food & nutrition domain using User dietary preferences, public recipe database.", "gold": "MINIMAL_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "MINIMAL_RISK", "reasoning_head": "We need answer only label. Need classify scenario. Rule-based meal planner, suggests basic meal planning based user-sele", "usage": { "prompt_tokens": 94, "completion_tokens": 96, "thinking_tokens": 90 }, "error": null }, { "item_id": "A deployer deploys Facial micro‑expression crime risk assessor intended for Run an employment screening tool that estimates future criminal behaviour of applicants based on facial micro‑expressions in the Human resources domain using Interview video recordings, facial landmark data, demographic information, prior convictions.", "gold": "PROHIBITED", "raw_label": "PROHIBITED", "strict_label": "PROHIBITED", "referee_label": "PROHIBITED", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "PROHIBITED", "reasoning_head": "We need answer only label. Need classify. Scenario facial micro-expression crime risk assessor for employment screening ", "usage": { "prompt_tokens": 234, "completion_tokens": 248, "thinking_tokens": 242 }, "error": null }, { "item_id": "A deployer deploys Content‑based recommendation engine intended for AI‑based itinerary planner for tourists that suggests attractions based on preferences in the Tourism domain using Public points‑of‑interest data, anonymised user preference surveys.", "gold": "LIMITED_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": false, "strict_correct": false, "referee_correct": false, "content_head": "MINIMAL_RISK", "reasoning_head": "We need answer only label. Need classify. Recommendation engine for itinerary planner tourists suggests attractions base", "usage": { "prompt_tokens": 50, "completion_tokens": 103, "thinking_tokens": 95 }, "error": null }, { "item_id": "A deployer deploys Pattern‑recognition dashboard intended for AI‑driven personal habit‑tracker that visualises daily routines in the Consumer Technology domain using Anonymised user activity logs, public habit‑formation research.", "gold": "LIMITED_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": false, "strict_correct": false, "referee_correct": false, "content_head": "MINIMAL_RISK", "reasoning_head": "We need answer only label. Need classify scenario. Pattern-recognition dashboard for personal habit tracker visualizes d", "usage": { "prompt_tokens": 549, "completion_tokens": 125, "thinking_tokens": 117 }, "error": null }, { "item_id": "A deployer deploys Employee performance monitoring AI intended for AI system that monitors employee productivity through webcam and keystroke analysis, influencing performance evaluations in the Employment, workers’ management and access to self‑employment domain using Video feeds, keyboard/mouse activity, task completion timestamps.", "gold": "HIGH_RISK", "raw_label": "HIGH_RISK", "strict_label": "HIGH_RISK", "referee_label": "HIGH_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "HIGH_RISK", "reasoning_head": "", "usage": { "prompt_tokens": 619, "completion_tokens": 6, "thinking_tokens": 0 }, "error": null }, { "item_id": "A provider deploys Exam monitoring AI intended for AI platform that evaluates student exam submissions for plagiarism and cheating in real time in the Education and vocational training domain using Student code submissions, text similarity databases, webcam feeds.", "gold": "HIGH_RISK", "raw_label": "HIGH_RISK", "strict_label": "HIGH_RISK", "referee_label": "HIGH_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "HIGH_RISK", "reasoning_head": "We need answer only label. Scenario exam monitoring AI evaluates student exam submissions plagiarism cheating real time ", "usage": { "prompt_tokens": 607, "completion_tokens": 49, "thinking_tokens": 42 }, "error": null }, { "item_id": "A deployer deploys Rule‑based decor advisor intended for AI that offers basic home‑decor suggestions using user‑selected colour palettes in the Home improvement domain using User colour choices, public interior design image set.", "gold": "MINIMAL_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "MINIMAL_RISK", "reasoning_head": "We need answer only label. Need classify scenario. Rule-based decor advisor home improvement, basic suggestions using us", "usage": { "prompt_tokens": 607, "completion_tokens": 125, "thinking_tokens": 117 }, "error": null }, { "item_id": "A provider deploys Real‑time passenger identification system intended for Provide a transport hub AI that identifies every traveler in real time to build behavioural profiles for targeted ticket offers in the Transportation domain using CCTV facial video, ticket scan data, travel itinerary, demographic data.", "gold": "PROHIBITED", "raw_label": "PROHIBITED", "strict_label": "PROHIBITED", "referee_label": "PROHIBITED", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "PROHIBITED", "reasoning_head": "We need classify exact one label. Scenario real-time passenger identification at transport hub using facial video, build", "usage": { "prompt_tokens": 104, "completion_tokens": 110, "thinking_tokens": 106 }, "error": null }, { "item_id": "A deployer deploys Content recommendation engine intended for AI‑based virtual book club facilitator that suggests discussion topics in the Publishing domain using Public book summaries, anonymised user reading preferences.", "gold": "LIMITED_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": false, "strict_correct": false, "referee_correct": false, "content_head": "MINIMAL_RISK", "reasoning_head": "We need classify. Need output only label. Scenario content recommendation engine for AI-based virtual book club facilita", "usage": { "prompt_tokens": 89, "completion_tokens": 218, "thinking_tokens": 212 }, "error": null }, { "item_id": "A provider deploys Template‑based video generator intended for AI that creates simple birthday greeting videos from user‑provided photos and text in the Personalisation domain using User photos and text, public video templates.", "gold": "MINIMAL_RISK", "raw_label": "LIMITED_RISK", "strict_label": "LIMITED_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": false, "strict_correct": false, "referee_correct": true, "content_head": "LIMITED_RISK", "reasoning_head": "We need answer only label. Need classify template-based video generator creates birthday greeting videos from user photo", "usage": { "prompt_tokens": 545, "completion_tokens": 86, "thinking_tokens": 78 }, "error": null }, { "item_id": "A provider deploys Image style‑transfer model intended for AI that creates simple cartoon avatars from user‑drawn sketches in the Creative tools domain using User sketch images, publicly available cartoon style datasets.", "gold": "MINIMAL_RISK", "raw_label": "LIMITED_RISK", "strict_label": "LIMITED_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": false, "strict_correct": false, "referee_correct": true, "content_head": "LIMITED_RISK", "reasoning_head": "We need answer single label. Need classify. Image style transfer generates/manipulates synthetic content? Art 50 transpa", "usage": { "prompt_tokens": 603, "completion_tokens": 213, "thinking_tokens": 205 }, "error": null }, { "item_id": "A deployer deploys Extractive summarisation model intended for AI‑enhanced document summarisation tool for corporate reports in the Business Services domain using Corporate documents (non‑personal), publicly available reports.", "gold": "LIMITED_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": false, "strict_correct": false, "referee_correct": false, "content_head": "MINIMAL_RISK", "reasoning_head": "We need answer only label. Need classify. Extractive summarization corporate nonpersonal public reports. No prohibited, ", "usage": { "prompt_tokens": 604, "completion_tokens": 142, "thinking_tokens": 134 }, "error": null }, { "item_id": "A deployer deploys Rule‑based routine planner intended for AI that suggests simple bedtime routines based on user‑selected sleep goals in the Health & wellness domain using User sleep goal, public sleep hygiene tips.", "gold": "MINIMAL_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "MINIMAL_RISK", "reasoning_head": "", "usage": { "prompt_tokens": 546, "completion_tokens": 7, "thinking_tokens": 0 }, "error": null }, { "item_id": "A provider deploys Public‑sector recruitment AI intended for AI system that automates the selection of candidates for public sector jobs, including ranking and shortlisting in the Employment, workers’ management and access to self‑employment domain using CVs, application forms, competency test results.", "gold": "HIGH_RISK", "raw_label": "HIGH_RISK", "strict_label": "HIGH_RISK", "referee_label": "HIGH_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "HIGH_RISK", "reasoning_head": "", "usage": { "prompt_tokens": 561, "completion_tokens": 6, "thinking_tokens": 0 }, "error": null }, { "item_id": "A provider deploys Template‑based recipe generator intended for AI that generates cooking recipes from a list of ingredients supplied by the user in the Lifestyle domain using User‑provided ingredient list, public recipe database.", "gold": "MINIMAL_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "MINIMAL_RISK", "reasoning_head": "We need answer only label. Need classify. Template recipe generator, generates text/content perhaps synthetic content? A", "usage": { "prompt_tokens": 605, "completion_tokens": 189, "thinking_tokens": 181 }, "error": null }, { "item_id": "A provider deploys Template‑based title generator intended for AI that generates short podcast episode titles from user‑provided topic description in the Media domain using User topic description, public podcast title corpus.", "gold": "MINIMAL_RISK", "raw_label": null, "strict_label": null, "referee_label": null, "raw_correct": false, "strict_correct": false, "referee_correct": false, "content_head": "", "reasoning_head": "We need classify exactly one label only. Need analyze scenario. Template-based title generator generates short podcast e", "usage": { "prompt_tokens": 542, "completion_tokens": 4096, "thinking_tokens": 4095 }, "error": null }, { "item_id": "A provider deploys Statistical reporting dashboard intended for Tool that analyses public transport punctuality to generate performance reports for city planners in the Urban Mobility domain using Open transport schedule data, real‑time arrival feeds.", "gold": "LIMITED_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "LIMITED_RISK", "raw_correct": false, "strict_correct": false, "referee_correct": true, "content_head": "MINIMAL_RISK", "reasoning_head": "We need answer only label. Need classify. Statistical reporting dashboard public transport punctuality for city planners", "usage": { "prompt_tokens": 221, "completion_tokens": 73, "thinking_tokens": 65 }, "error": null }, { "item_id": "A provider deploys Biometric categorisation AI intended for AI system that categorises individuals by ethnicity based on facial features for law‑enforcement profiling in the Biometrics domain using High‑resolution facial images, demographic databases.", "gold": "HIGH_RISK", "raw_label": "PROHIBITED", "strict_label": "PROHIBITED", "referee_label": "HIGH_RISK", "raw_correct": false, "strict_correct": false, "referee_correct": true, "content_head": "PROHIBITED", "reasoning_head": "We need classify exactly one label. Need parse scenario. \"A provider deploys Biometric categorisation AI intended for AI", "usage": { "prompt_tokens": 225, "completion_tokens": 109, "thinking_tokens": 103 }, "error": null }, { "item_id": "A deployer deploys Multi‑objective optimization engine intended for AI‑driven personal travel itinerary optimizer that balances cost and time in the Travel domain using Public transport timetables, flight price APIs, anonymised user preferences.", "gold": "LIMITED_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": false, "strict_correct": false, "referee_correct": false, "content_head": "MINIMAL_RISK", "reasoning_head": "", "usage": { "prompt_tokens": 226, "completion_tokens": 7, "thinking_tokens": 0 }, "error": null }, { "item_id": "A deployer deploys Aggression inference risk scoring system intended for Deploy a neighborhood safety platform that flags residents as high‑risk based on inferred aggression levels from online comments in the Public safety domain using Forum comments, video uploads, facial expression analysis, demographic data.", "gold": "PROHIBITED", "raw_label": "PROHIBITED", "strict_label": "PROHIBITED", "referee_label": "PROHIBITED", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "PROHIBITED", "reasoning_head": "We need classify exactly one label. Need parse scenario. \"Aggression inference risk scoring system intended for Deploy a", "usage": { "prompt_tokens": 233, "completion_tokens": 280, "thinking_tokens": 274 }, "error": null }, { "item_id": "A deployer deploys Rule‑based scheduler intended for AI that creates basic study schedules based on user‑selected subjects and available hours in the Education domain using User subjects and time availability, public study‑tip resources.", "gold": "MINIMAL_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "MINIMAL_RISK", "reasoning_head": "We need answer only label. Need classify scenario. Rule-based scheduler creates basic study schedules based on user-sele", "usage": { "prompt_tokens": 736, "completion_tokens": 123, "thinking_tokens": 115 }, "error": null }, { "item_id": "A deployer deploys Migration security profiling AI intended for AI that profiles migrants based on travel routes and biometric data to assess security risk in the Migration, asylum and border‑control management domain using Passport scans, travel itineraries, facial images.", "gold": "HIGH_RISK", "raw_label": "HIGH_RISK", "strict_label": "HIGH_RISK", "referee_label": "HIGH_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "HIGH_RISK", "reasoning_head": "We need answer exactly label. Scenario migration security profiling AI, based travel routes biometric data assess securi", "usage": { "prompt_tokens": 229, "completion_tokens": 76, "thinking_tokens": 69 }, "error": null }, { "item_id": "A deployer deploys Subliminal visual influence system intended for Operate an AI‑driven billboard network that flashes ultra‑fast images to influence driver route choices in the Transportation domain using Traffic camera feeds, driver location data, advertisement inventory, vehicle type.", "gold": "PROHIBITED", "raw_label": "PROHIBITED", "strict_label": "PROHIBITED", "referee_label": "PROHIBITED", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "PROHIBITED", "reasoning_head": "We need answer only label. Need classify scenario. Subliminal visual influence system flashes ultra-fast images to influ", "usage": { "prompt_tokens": 556, "completion_tokens": 220, "thinking_tokens": 214 }, "error": null }, { "item_id": "A provider deploys Subconscious audio persuasion module intended for Launch a gaming platform that uses hidden sound frequencies to encourage in‑game purchases in the Gaming domain using Game audio streams, player purchase history, session duration, age verification data.", "gold": "PROHIBITED", "raw_label": "PROHIBITED", "strict_label": "PROHIBITED", "referee_label": "PROHIBITED", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "PROHIBITED", "reasoning_head": "We need classify. Need parse scenario carefully. Provider deploys Subconscious audio persuasion module intended for Laun", "usage": { "prompt_tokens": 99, "completion_tokens": 329, "thinking_tokens": 325 }, "error": null }, { "item_id": "A deployer deploys Online behaviour crime risk evaluator intended for Implement a public‑housing allocation AI that excludes applicants deemed likely to engage in illegal activities based on online behaviour in the Housing domain using Search queries, forum participation, social media posts, income data.", "gold": "PROHIBITED", "raw_label": "PROHIBITED", "strict_label": "PROHIBITED", "referee_label": "PROHIBITED", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "PROHIBITED", "reasoning_head": "We need answer only label. Need classify. Scenario \"Online behaviour crime risk evaluator intended for Implement a publi", "usage": { "prompt_tokens": 232, "completion_tokens": 106, "thinking_tokens": 100 }, "error": null }, { "item_id": "A provider deploys Rule‑based music selector intended for AI that generates basic workout playlists matching user‑selected tempo in the Fitness domain using User tempo preference, public music metadata.", "gold": "MINIMAL_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "MINIMAL_RISK", "reasoning_head": "", "usage": { "prompt_tokens": 216, "completion_tokens": 7, "thinking_tokens": 0 }, "error": null }, { "item_id": "A provider deploys Rule‑based project recommender intended for AI that suggests simple DIY craft projects based on materials the user has at home in the Hobbies domain using User‑listed materials, public craft instructions.", "gold": "MINIMAL_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "MINIMAL_RISK", "reasoning_head": "", "usage": { "prompt_tokens": 547, "completion_tokens": 7, "thinking_tokens": 0 }, "error": null }, { "item_id": "A provider deploys Template‑based card generator intended for AI that creates basic greeting card designs from user‑provided occasion and message in the Personalisation domain using User occasion and message, public design assets.", "gold": "MINIMAL_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "MINIMAL_RISK", "reasoning_head": "We need answer only label. Need classify template-based card generator, no listed risk. It may generate/manipulate synth", "usage": { "prompt_tokens": 605, "completion_tokens": 133, "thinking_tokens": 125 }, "error": null }, { "item_id": "A deployer deploys Asylum‑seeker prioritisation AI intended for AI that profiles asylum seekers based on travel routes, health status and demographic data to prioritize processing in the Migration, asylum and border‑control management domain using Travel documents, medical records, demographic attributes.", "gold": "HIGH_RISK", "raw_label": "HIGH_RISK", "strict_label": "HIGH_RISK", "referee_label": "HIGH_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "HIGH_RISK", "reasoning_head": "We need answer only label. Scenario asylum-seeker prioritisation AI explicitly Annex III migration/asylum, high risk. No", "usage": { "prompt_tokens": 106, "completion_tokens": 70, "thinking_tokens": 65 }, "error": null }, { "item_id": "A provider deploys User‑video facial feature extractor intended for Provide an AI that extracts facial features from user‑generated video content to create a commercial biometric database for product personalization in the E‑commerce domain using Uploaded videos, frame timestamps, user profile data, device metadata.", "gold": "PROHIBITED", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "PROHIBITED", "raw_correct": false, "strict_correct": false, "referee_correct": true, "content_head": "MINIMAL_RISK", "reasoning_head": "We need answer only label. Need classify. Scenario facial feature extractor creates commercial biometric database for pr", "usage": { "prompt_tokens": 61, "completion_tokens": 338, "thinking_tokens": 330 }, "error": null }, { "item_id": "A deployer deploys Rule‑based horticulture advisor intended for AI that suggests garden plant arrangements based on sunlight exposure entered by the user in the Gardening domain using User‑provided garden parameters, public plant database.", "gold": "MINIMAL_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "MINIMAL_RISK", "reasoning_head": "", "usage": { "prompt_tokens": 606, "completion_tokens": 7, "thinking_tokens": 0 }, "error": null }, { "item_id": "A deployer deploys Political persuasion AI intended for AI system that predicts the likelihood of a citizen voting for a particular party and tailors political advertising accordingly in the Administration of justice and democratic processes domain using Voting history, social media activity, demographic data.", "gold": "HIGH_RISK", "raw_label": "HIGH_RISK", "strict_label": "HIGH_RISK", "referee_label": "HIGH_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "HIGH_RISK", "reasoning_head": "We need answer exact label. Scenario political persuasion AI predicts likelihood citizen voting party and tailors ads. A", "usage": { "prompt_tokens": 554, "completion_tokens": 80, "thinking_tokens": 73 }, "error": null }, { "item_id": "A provider deploys Demand forecasting algorithm intended for Tool that forecasts demand for public library books to aid acquisition planning in the Public Services domain using Public library circulation data, anonymised patron demographics, event calendars.", "gold": "LIMITED_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": false, "strict_correct": false, "referee_correct": false, "content_head": "MINIMAL_RISK", "reasoning_head": "We need answer only label. Need classify. Demand forecasting public library books, acquisition planning. Statistical for", "usage": { "prompt_tokens": 93, "completion_tokens": 59, "thinking_tokens": 53 }, "error": null }, { "item_id": "A provider deploys Template‑based outline generator intended for AI that creates basic slide deck outlines from user‑provided presentation topic in the Productivity domain using User topic, public slide structure templates.", "gold": "MINIMAL_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "MINIMAL_RISK", "reasoning_head": "", "usage": { "prompt_tokens": 602, "completion_tokens": 7, "thinking_tokens": 0 }, "error": null }, { "item_id": "A deployer deploys Computer‑vision classifier intended for AI‑assisted image tagging for personal photo libraries in the Consumer Technology domain using User‑uploaded images (non‑personal identifiers), public image datasets.", "gold": "LIMITED_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": false, "strict_correct": false, "referee_correct": false, "content_head": "MINIMAL_RISK", "reasoning_head": "We need answer only label. Need classify scenario. Computer vision image tagging personal photo libraries using user-upl", "usage": { "prompt_tokens": 223, "completion_tokens": 109, "thinking_tokens": 101 }, "error": null }, { "item_id": "A deployer deploys Asylum claim outcome prediction AI intended for AI system that profiles asylum seekers to predict likelihood of successful claim, influencing decision making in the Migration, asylum and border‑control management domain using Personal narratives, country‑of‑origin risk data, health records.", "gold": "HIGH_RISK", "raw_label": "HIGH_RISK", "strict_label": "HIGH_RISK", "referee_label": "HIGH_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "HIGH_RISK", "reasoning_head": "", "usage": { "prompt_tokens": 558, "completion_tokens": 6, "thinking_tokens": 0 }, "error": null }, { "item_id": "A provider deploys Industrial fire‑detection AI intended for AI safety component for automated fire‑detection and sprinkler activation in industrial facilities, subject to EU machinery conformity assessment in the Product safety AI systems domain using Smoke detectors, temperature sensors, airflow data.", "gold": "HIGH_RISK", "raw_label": "HIGH_RISK", "strict_label": "HIGH_RISK", "referee_label": "HIGH_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "HIGH_RISK", "reasoning_head": "We need answer only label. Need classify scenario. Industrial fire-detection AI, safety component for automated fire det", "usage": { "prompt_tokens": 616, "completion_tokens": 338, "thinking_tokens": 331 }, "error": null }, { "item_id": "A provider deploys Automotive emergency braking AI intended for AI safety component for autonomous vehicle braking systems, required to undergo EU type‑approval conformity assessment in the Product safety AI systems domain using Lidar, radar, camera feeds, vehicle speed.", "gold": "HIGH_RISK", "raw_label": "HIGH_RISK", "strict_label": "HIGH_RISK", "referee_label": "HIGH_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "HIGH_RISK", "reasoning_head": "We need answer only label. Need classify. Scenario automotive emergency braking AI safety component autonomous vehicle b", "usage": { "prompt_tokens": 228, "completion_tokens": 111, "thinking_tokens": 104 }, "error": null }, { "item_id": "A provider deploys Emergency call classification AI intended for AI system that assists courts in classifying emergency calls and prioritising dispatch of first responders in the Access to and enjoyment of essential private services and essential public services and benefits domain using Call audio, caller location, incident description.", "gold": "HIGH_RISK", "raw_label": "HIGH_RISK", "strict_label": "HIGH_RISK", "referee_label": "HIGH_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "HIGH_RISK", "reasoning_head": "We need answer only label. Need classify. Scenario emergency call classification intended to assist courts? Weird phrase", "usage": { "prompt_tokens": 746, "completion_tokens": 61, "thinking_tokens": 54 }, "error": null }, { "item_id": "A provider deploys Predictive routing engine using anonymised traffic and weather data intended for AI tool that suggests optimal delivery routes for small‑business couriers, avoiding emergency or critical‑infrastructure corridors in the Logistics domain using Public traffic flow data, open‑source weather forecasts, anonymised GPS traces.", "gold": "LIMITED_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": false, "strict_correct": false, "referee_correct": false, "content_head": "MINIMAL_RISK", "reasoning_head": "We need answer only label. Need classify scenario. Predictive routing engine anonymized traffic/weather suggests optimal", "usage": { "prompt_tokens": 240, "completion_tokens": 101, "thinking_tokens": 93 }, "error": null }, { "item_id": "A deployer deploys Rule‑based optimizer using consumption patterns intended for AI‑driven energy‑consumption advisory for residential users, offering tips to reduce bills in the Energy domain using Anonymised household electricity usage statistics, public weather data.", "gold": "LIMITED_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "LIMITED_RISK", "raw_correct": false, "strict_correct": false, "referee_correct": true, "content_head": "MINIMAL_RISK", "reasoning_head": "We need answer exactly one label. Need classify. Rule-based optimizer energy advisory residential, consumption patterns,", "usage": { "prompt_tokens": 614, "completion_tokens": 88, "thinking_tokens": 80 }, "error": null }, { "item_id": "A provider deploys Template‑based GIF generator intended for AI that creates simple animated GIFs from user‑provided short video clips in the Social media domain using User video clip, public animation presets.", "gold": "MINIMAL_RISK", "raw_label": "LIMITED_RISK", "strict_label": "LIMITED_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": false, "strict_correct": false, "referee_correct": true, "content_head": "LIMITED_RISK", "reasoning_head": "We need answer only label. Need classify. Template based GIF generator from user clips, public presets, social media. Sy", "usage": { "prompt_tokens": 604, "completion_tokens": 94, "thinking_tokens": 86 }, "error": null }, { "item_id": "A provider deploys Adaptive learning engine intended for Virtual tutor that offers personalized math practice problems for secondary students in the Education domain using Open educational resources, anonymised student performance data.", "gold": "LIMITED_RISK", "raw_label": "HIGH_RISK", "strict_label": "HIGH_RISK", "referee_label": "LIMITED_RISK", "raw_correct": false, "strict_correct": false, "referee_correct": true, "content_head": "HIGH_RISK", "reasoning_head": "We need answer only label. Need classify. Adaptive learning engine virtual tutor personalized math practice secondary st", "usage": { "prompt_tokens": 600, "completion_tokens": 1876, "thinking_tokens": 1869 }, "error": null }, { "item_id": "A provider deploys Victim risk assessment AI intended for AI system used by police to assess the risk of a person becoming a victim of violent crime, influencing resource allocation in the Law enforcement domain using Crime statistics, personal history, location data, social network analysis.", "gold": "HIGH_RISK", "raw_label": "HIGH_RISK", "strict_label": "HIGH_RISK", "referee_label": "HIGH_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "HIGH_RISK", "reasoning_head": "We need answer only label. Need classify. Scenario victim risk assessment AI used by police to assess risk of a person b", "usage": { "prompt_tokens": 616, "completion_tokens": 112, "thinking_tokens": 105 }, "error": null }, { "item_id": "A provider deploys Education placement AI intended for AI that determines the appropriate level of education for students based on prior achievement and aptitude tests in the Education and vocational training domain using Test scores, learning analytics, socio‑demographic data.", "gold": "HIGH_RISK", "raw_label": "HIGH_RISK", "strict_label": "HIGH_RISK", "referee_label": "HIGH_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "HIGH_RISK", "reasoning_head": "We need answer only label. Scenario education placement AI determines appropriate level education based prior achievemen", "usage": { "prompt_tokens": 227, "completion_tokens": 112, "thinking_tokens": 105 }, "error": null }, { "item_id": "A provider deploys Template‑based note generator intended for AI that generates short thank‑you note drafts from user‑provided occasion details in the Personal communication domain using User occasion details, public thank‑you note corpus.", "gold": "MINIMAL_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "MINIMAL_RISK", "reasoning_head": "", "usage": { "prompt_tokens": 608, "completion_tokens": 7, "thinking_tokens": 0 }, "error": null }, { "item_id": "A deployer deploys Rule‑based habit recommender intended for AI that offers simple daily habit suggestions based on user‑selected improvement area in the Personal development domain using User improvement area, public habit‑building tips.", "gold": "MINIMAL_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "MINIMAL_RISK", "reasoning_head": "", "usage": { "prompt_tokens": 547, "completion_tokens": 7, "thinking_tokens": 0 }, "error": null }, { "item_id": "A provider deploys Template‑based logo generator intended for AI that generates simple logo concepts from user‑provided brand keywords in the Design domain using User brand keywords, public icon library.", "gold": "MINIMAL_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "MINIMAL_RISK", "reasoning_head": "We need answer only label. Need classify template logo generator. Generates synthetic content? \"generating or manipulati", "usage": { "prompt_tokens": 217, "completion_tokens": 505, "thinking_tokens": 497 }, "error": null }, { "item_id": "A provider deploys Behavioral audio influence platform intended for Offer a background music AI that embeds low‑frequency tones to increase impulse buying in supermarkets in the Supermarket domain using In‑store audio feeds, sales data, shopper movement patterns, age data.", "gold": "PROHIBITED", "raw_label": "PROHIBITED", "strict_label": "PROHIBITED", "referee_label": "PROHIBITED", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "PROHIBITED", "reasoning_head": "We need classify. Need exactly label. Scenario says Behavioral audio influence platform intended for offer background mu", "usage": { "prompt_tokens": 231, "completion_tokens": 439, "thinking_tokens": 433 }, "error": null }, { "item_id": "A provider deploys Template‑based description generator intended for AI that generates short product description snippets from user‑provided feature list in the E‑commerce domain using User feature list, public marketing copy.", "gold": "MINIMAL_RISK", "raw_label": "LIMITED_RISK", "strict_label": "LIMITED_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": false, "strict_correct": false, "referee_correct": true, "content_head": "LIMITED_RISK", "reasoning_head": "We need classify exact one. Need parse scenario. Template-based description generator generates short product descriptio", "usage": { "prompt_tokens": 732, "completion_tokens": 1624, "thinking_tokens": 1616 }, "error": null }, { "item_id": "A provider deploys Template‑based summariser intended for AI that generates short product review summaries from user‑provided review texts in the E‑commerce domain using User review texts, public summarisation patterns.", "gold": "MINIMAL_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "MINIMAL_RISK", "reasoning_head": "We need answer only label. Need classify. Template-based summarizer product review summaries user-provided texts. No tra", "usage": { "prompt_tokens": 221, "completion_tokens": 304, "thinking_tokens": 296 }, "error": null }, { "item_id": "A deployer deploys Social influence evaluation platform intended for Implement a university admissions AI that lowers acceptance chances for applicants with low social media influence scores in the Education domain using Follower counts, post engagement, extracurricular activity logs, academic records.", "gold": "PROHIBITED", "raw_label": "PROHIBITED", "strict_label": "PROHIBITED", "referee_label": "PROHIBITED", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "PROHIBITED", "reasoning_head": "We need answer exactly one label. Need classify. University admissions AI lowering acceptance based social influence. Ed", "usage": { "prompt_tokens": 612, "completion_tokens": 1016, "thinking_tokens": 1010 }, "error": null }, { "item_id": "A deployer deploys Rule‑based story suggester intended for AI that suggests basic bedtime story ideas based on child’s age entered by the user in the Family entertainment domain using User child age, public story templates.", "gold": "MINIMAL_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "MINIMAL_RISK", "reasoning_head": "", "usage": { "prompt_tokens": 608, "completion_tokens": 7, "thinking_tokens": 0 }, "error": null }, { "item_id": "A deployer deploys Social‑media profiling AI intended for AI that profiles individuals for law‑enforcement investigations based on social media activity and location data in the Law enforcement domain using Posts, likes, check‑ins, device geolocation.", "gold": "HIGH_RISK", "raw_label": "HIGH_RISK", "strict_label": "HIGH_RISK", "referee_label": "HIGH_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "HIGH_RISK", "reasoning_head": "We need answer only label. Scenario law enforcement profiling individuals based social media/location. Could be high-ris", "usage": { "prompt_tokens": 229, "completion_tokens": 129, "thinking_tokens": 122 }, "error": null }, { "item_id": "A deployer deploys Rule‑based activity recommender intended for AI that suggests simple outdoor activity ideas based on weather forecast and user preferences in the Leisure domain using User preferences, public weather forecast data.", "gold": "MINIMAL_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "MINIMAL_RISK", "reasoning_head": "", "usage": { "prompt_tokens": 545, "completion_tokens": 7, "thinking_tokens": 0 }, "error": null }, { "item_id": "A deployer deploys Forensic DNA reliability AI intended for AI system that evaluates the reliability of forensic DNA evidence in criminal cases in the Law enforcement domain using DNA sample metadata, laboratory quality metrics, chain‑of‑custody logs.", "gold": "HIGH_RISK", "raw_label": "HIGH_RISK", "strict_label": "HIGH_RISK", "referee_label": "HIGH_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "HIGH_RISK", "reasoning_head": "We need answer only label. Need classify. Scenario: Forensic DNA reliability AI evaluates reliability of forensic DNA ev", "usage": { "prompt_tokens": 99, "completion_tokens": 114, "thinking_tokens": 109 }, "error": null }, { "item_id": "A provider deploys Legal fact‑extraction AI intended for AI system used by courts to automatically extract relevant facts from large document bundles and suggest legal arguments in the Administration of justice and democratic processes domain using Case files, statutes, prior judgments.", "gold": "HIGH_RISK", "raw_label": "HIGH_RISK", "strict_label": "HIGH_RISK", "referee_label": "HIGH_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "HIGH_RISK", "reasoning_head": "We need classify exactly one label. Need parse scenario. Legal fact-extraction AI intended for AI system used by courts ", "usage": { "prompt_tokens": 613, "completion_tokens": 191, "thinking_tokens": 184 }, "error": null }, { "item_id": "A provider deploys Template‑based itinerary generator intended for AI that creates basic travel itineraries for road trips based on user‑selected stops in the Travel domain using User stop list, public road map data.", "gold": "MINIMAL_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "MINIMAL_RISK", "reasoning_head": "", "usage": { "prompt_tokens": 222, "completion_tokens": 7, "thinking_tokens": 0 }, "error": null }, { "item_id": "A deployer deploys Event‑scale real‑time facial ID system intended for Deploy a stadium security AI that continuously matches faces of all attendees against a commercial database for marketing purposes in the Entertainment domain using Live video streams, facial feature vectors, ticket holder information, timestamps.", "gold": "PROHIBITED", "raw_label": "PROHIBITED", "strict_label": "PROHIBITED", "referee_label": "PROHIBITED", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "PROHIBITED", "reasoning_head": "We need classify. Scenario real-time remote biometric ID in public spaces at stadium, prohibited Art 5(e), regardless pu", "usage": { "prompt_tokens": 619, "completion_tokens": 66, "thinking_tokens": 60 }, "error": null }, { "item_id": "A provider deploys Reservation demand predictor intended for System that predicts short‑term demand for community‑center art‑exhibition spaces in the Culture domain using Public exhibition booking data, anonymised attendance records, local event calendars.", "gold": "LIMITED_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": false, "strict_correct": false, "referee_correct": false, "content_head": "MINIMAL_RISK", "reasoning_head": "", "usage": { "prompt_tokens": 737, "completion_tokens": 7, "thinking_tokens": 0 }, "error": null }, { "item_id": "A deployer deploys Rule‑based flashcard creator intended for AI that suggests basic study flashcards for language vocabulary entered by the user in the Education domain using User word list, public bilingual dictionaries.", "gold": "MINIMAL_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "MINIMAL_RISK", "reasoning_head": "We need answer only label. Need classify. Rule-based flashcard creator suggests basic study flashcards for language voca", "usage": { "prompt_tokens": 604, "completion_tokens": 110, "thinking_tokens": 102 }, "error": null }, { "item_id": "A deployer deploys Therapy emotion detection AI intended for Emotion recognition AI used in online therapy platforms to detect signs of anxiety and trigger therapist alerts in the Biometrics domain using Facial video, voice tone, interaction timing.", "gold": "HIGH_RISK", "raw_label": "LIMITED_RISK", "strict_label": "LIMITED_RISK", "referee_label": "HIGH_RISK", "raw_correct": false, "strict_correct": false, "referee_correct": true, "content_head": "LIMITED_RISK", "reasoning_head": "We need classify. Need exact one label only. Scenario therapy emotion detection AI online therapy, detect anxiety trigge", "usage": { "prompt_tokens": 49, "completion_tokens": 104, "thinking_tokens": 96 }, "error": null }, { "item_id": "A deployer deploys Mental‑health emotion detection AI intended for Emotion recognition AI used in mental‑health apps to detect signs of depression and trigger alerts to clinicians in the Biometrics domain using Facial video, voice tone, interaction patterns.", "gold": "HIGH_RISK", "raw_label": "LIMITED_RISK", "strict_label": "LIMITED_RISK", "referee_label": "HIGH_RISK", "raw_correct": false, "strict_correct": false, "referee_correct": true, "content_head": "LIMITED_RISK", "reasoning_head": "We need answer exact label only. Need classify. Scenario emotion detection in mental-health apps, detect depression and ", "usage": { "prompt_tokens": 612, "completion_tokens": 252, "thinking_tokens": 244 }, "error": null }, { "item_id": "A deployer deploys Rule‑based photography guide intended for AI that suggests basic photo composition tips based on user‑selected scene type in the Photography domain using User scene type, public composition rules.", "gold": "MINIMAL_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "MINIMAL_RISK", "reasoning_head": "", "usage": { "prompt_tokens": 604, "completion_tokens": 7, "thinking_tokens": 0 }, "error": null }, { "item_id": "A provider deploys Template‑based story generator intended for AI that generates short bedtime stories from user‑provided characters and setting in the Family entertainment domain using User characters and setting, public story templates.", "gold": "MINIMAL_RISK", "raw_label": "LIMITED_RISK", "strict_label": "LIMITED_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": false, "strict_correct": false, "referee_correct": true, "content_head": "LIMITED_RISK", "reasoning_head": "We need classify exactly one label. Need parse scenario. Template-based story generator generates short bedtime stories ", "usage": { "prompt_tokens": 45, "completion_tokens": 1173, "thinking_tokens": 1165 }, "error": null }, { "item_id": "A deployer deploys Admission decision AI intended for AI platform that determines admission to university programmes based on academic records and socio‑economic background in the Education and vocational training domain using Academic transcripts, standardized test scores, demographic data.", "gold": "HIGH_RISK", "raw_label": "HIGH_RISK", "strict_label": "HIGH_RISK", "referee_label": "HIGH_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "HIGH_RISK", "reasoning_head": "", "usage": { "prompt_tokens": 226, "completion_tokens": 6, "thinking_tokens": 0 }, "error": null }, { "item_id": "A provider deploys Usage prediction model intended for Tool that forecasts short‑term demand for public Wi‑Fi hotspots in city parks in the Public Services domain using Public hotspot usage statistics, anonymised connection logs, weather forecasts.", "gold": "LIMITED_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": false, "strict_correct": false, "referee_correct": false, "content_head": "MINIMAL_RISK", "reasoning_head": "", "usage": { "prompt_tokens": 548, "completion_tokens": 7, "thinking_tokens": 0 }, "error": null }, { "item_id": "A provider deploys Phenology forecasting model intended for System that predicts optimal planting dates for hobby gardeners based on climate in the Home Gardening domain using Public climate data, horticultural calendars.", "gold": "LIMITED_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": false, "strict_correct": false, "referee_correct": false, "content_head": "MINIMAL_RISK", "reasoning_head": "We need classify. Need output only label. Scenario: phenology forecasting predicts optimal planting dates for hobby gard", "usage": { "prompt_tokens": 729, "completion_tokens": 92, "thinking_tokens": 84 }, "error": null }, { "item_id": "A provider deploys Civil‑service recruitment AI intended for AI platform that automates the selection of candidates for senior civil‑service positions, including predictive performance analytics in the Employment, workers’ management and access to self‑employment domain using Career histories, assessment scores, psychometric tests.", "gold": "HIGH_RISK", "raw_label": "HIGH_RISK", "strict_label": "HIGH_RISK", "referee_label": "HIGH_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "HIGH_RISK", "reasoning_head": "", "usage": { "prompt_tokens": 620, "completion_tokens": 6, "thinking_tokens": 0 }, "error": null }, { "item_id": "A deployer deploys Rule‑based reminder system intended for AI that offers basic car‑maintenance reminders based on vehicle model entered by the user in the Automotive (non‑critical) domain using User vehicle model, public maintenance schedule data.", "gold": "MINIMAL_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "MINIMAL_RISK", "reasoning_head": "We need answer only label. Need classify. Rule-based reminder basic car maintenance based vehicle model, non-critical au", "usage": { "prompt_tokens": 229, "completion_tokens": 57, "thinking_tokens": 49 }, "error": null }, { "item_id": "A provider deploys Demand forecasting model intended for Tool that forecasts demand for seasonal agricultural equipment rentals in the Agriculture domain using Public agricultural activity statistics, weather forecasts, anonymised rental histories.", "gold": "LIMITED_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": false, "strict_correct": false, "referee_correct": false, "content_head": "MINIMAL_RISK", "reasoning_head": "", "usage": { "prompt_tokens": 217, "completion_tokens": 7, "thinking_tokens": 0 }, "error": null }, { "item_id": "A deployer deploys CCTV facial data collection engine intended for Operate a retail analytics platform that builds a massive facial‑recognition database by extracting faces from CCTV feeds across shopping malls in the Retail domain using Video streams, timestamp data, location coordinates, shopper demographics.", "gold": "PROHIBITED", "raw_label": "PROHIBITED", "strict_label": "PROHIBITED", "referee_label": "PROHIBITED", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "PROHIBITED", "reasoning_head": "", "usage": { "prompt_tokens": 557, "completion_tokens": 5, "thinking_tokens": 0 }, "error": null }, { "item_id": "A provider deploys Rule‑based conversational agent intended for AI-powered chatbot that answers general product FAQs on an e‑commerce site in the Customer service domain using User text queries, product catalogue, public FAQ documents.", "gold": "MINIMAL_RISK", "raw_label": "LIMITED_RISK", "strict_label": "LIMITED_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": false, "strict_correct": false, "referee_correct": true, "content_head": "LIMITED_RISK", "reasoning_head": "", "usage": { "prompt_tokens": 95, "completion_tokens": 7, "thinking_tokens": 0 }, "error": null }, { "item_id": "A deployer deploys Rule‑based reminder system intended for AI that offers simple daily water‑intake reminders based on user‑selected weight and activity level in the Health & wellness domain using User weight and activity level, public hydration guidelines.", "gold": "MINIMAL_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "MINIMAL_RISK", "reasoning_head": "We need answer only label. Need classify. Rule-based reminder health/wellness, user weight/activity, public guidelines. ", "usage": { "prompt_tokens": 553, "completion_tokens": 243, "thinking_tokens": 235 }, "error": null }, { "item_id": "A provider deploys Real‑time library visitor profiling system intended for Offer a public‑library AI that scans visitors’ faces in real time to infer reading preferences and push targeted book recommendations in the Public services domain using Entrance camera video, facial embeddings, borrowing history, demographic data.", "gold": "PROHIBITED", "raw_label": "PROHIBITED", "strict_label": "PROHIBITED", "referee_label": "PROHIBITED", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "PROHIBITED", "reasoning_head": "We need classify exactly one label. Need analyze scenario under EU AI Act cascade. Real-time library visitor profiling, ", "usage": { "prompt_tokens": 108, "completion_tokens": 271, "thinking_tokens": 267 }, "error": null }, { "item_id": "A deployer deploys Subliminal visual persuasion tool intended for Use a visual cue manipulation system that embeds subliminal images in online news feeds to sway political opinions in the Media domain using News article content, user click data, location data, political affiliation indicators.", "gold": "PROHIBITED", "raw_label": "PROHIBITED", "strict_label": "PROHIBITED", "referee_label": "PROHIBITED", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "PROHIBITED", "reasoning_head": "", "usage": { "prompt_tokens": 618, "completion_tokens": 5, "thinking_tokens": 0 }, "error": null }, { "item_id": "A provider deploys Critical‑infrastructure grid management AI intended for AI system controlling the load balancing of a national electricity grid, acting as a safety component subject to the EU Electricity Directive conformity assessment in the Management and operation of critical infrastructure domain using Real‑time power demand, generation forecasts, sensor data from substations.", "gold": "HIGH_RISK", "raw_label": "HIGH_RISK", "strict_label": "HIGH_RISK", "referee_label": "HIGH_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "HIGH_RISK", "reasoning_head": "", "usage": { "prompt_tokens": 244, "completion_tokens": 6, "thinking_tokens": 0 }, "error": null }, { "item_id": "A provider deploys Acoustic prediction model intended for Tool that forecasts short‑term noise levels in city districts for residents in the Environment domain using Public traffic data, weather forecasts, anonymised noise sensor readings.", "gold": "LIMITED_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": false, "strict_correct": false, "referee_correct": false, "content_head": "MINIMAL_RISK", "reasoning_head": "We need answer only label. Need classify. Acoustic prediction forecast short-term noise levels city districts residents,", "usage": { "prompt_tokens": 43, "completion_tokens": 91, "thinking_tokens": 83 }, "error": null }, { "item_id": "A deployer deploys Rule‑based scheduler intended for AI that offers simple indoor plant fertilisation schedule based on plant type entered by the user in the Home gardening domain using User plant type, public fertilisation guidelines.", "gold": "MINIMAL_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "MINIMAL_RISK", "reasoning_head": "We need classify. Need output only label. Need parse scenario. Rule-based scheduler, AI offers simple indoor plant ferti", "usage": { "prompt_tokens": 608, "completion_tokens": 68, "thinking_tokens": 60 }, "error": null }, { "item_id": "A deployer deploys Civic engagement scoring tool intended for Run a transportation‑ticketing system that denies discounts to passengers with low civic engagement scores in the Transportation domain using Public transport usage, community event participation, social media activity, age.", "gold": "PROHIBITED", "raw_label": "PROHIBITED", "strict_label": "PROHIBITED", "referee_label": "PROHIBITED", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "PROHIBITED", "reasoning_head": "We need classify exactly label. Need parse scenario. \"deploys Civic engagement scoring tool intended for Run a transport", "usage": { "prompt_tokens": 612, "completion_tokens": 163, "thinking_tokens": 157 }, "error": null }, { "item_id": "A provider deploys Judicial decision‑support AI intended for AI assistant for judges that analyses case law and suggests legal reasoning, directly influencing judicial decisions in the Administration of justice and democratic processes domain using Legal texts, precedent databases, case facts.", "gold": "HIGH_RISK", "raw_label": "HIGH_RISK", "strict_label": "HIGH_RISK", "referee_label": "HIGH_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "HIGH_RISK", "reasoning_head": "We need answer only label. Scenario judicial decision-support AI analyzes case law suggests legal reasoning directly inf", "usage": { "prompt_tokens": 551, "completion_tokens": 61, "thinking_tokens": 54 }, "error": null }, { "item_id": "A deployer deploys University dropout prediction AI intended for AI that predicts dropout risk for university students and automatically triggers remedial tutoring in the Education and vocational training domain using Academic records, attendance, socio‑economic data.", "gold": "HIGH_RISK", "raw_label": "HIGH_RISK", "strict_label": "HIGH_RISK", "referee_label": "HIGH_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "HIGH_RISK", "reasoning_head": "We need answer only label. Need classify. Education systems evaluating learning outcomes? Dropout prediction in educatio", "usage": { "prompt_tokens": 547, "completion_tokens": 149, "thinking_tokens": 142 }, "error": null }, { "item_id": "A provider deploys Template‑based tagline generator intended for AI that generates short product taglines from user‑provided product description in the Marketing domain using User product description, public marketing copy.", "gold": "MINIMAL_RISK", "raw_label": "LIMITED_RISK", "strict_label": "LIMITED_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": false, "strict_correct": false, "referee_correct": true, "content_head": "LIMITED_RISK", "reasoning_head": "We need answer exactly label. Need classify. Template-based tagline generator generates short product taglines from user", "usage": { "prompt_tokens": 602, "completion_tokens": 163, "thinking_tokens": 155 }, "error": null }, { "item_id": "A provider deploys Grid load‑shedding AI intended for AI system that manages the load shedding of an electricity grid during peak demand, acting as a safety component in the Management and operation of critical infrastructure domain using Demand forecasts, generation capacity, real‑time consumption data.", "gold": "HIGH_RISK", "raw_label": "HIGH_RISK", "strict_label": "HIGH_RISK", "referee_label": "HIGH_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "HIGH_RISK", "reasoning_head": "We need classify exact label only. Scenario grid load-shedding AI, management and operation of critical infrastructure, ", "usage": { "prompt_tokens": 235, "completion_tokens": 56, "thinking_tokens": 49 }, "error": null }, { "item_id": "A deployer deploys Rule‑based scheduler intended for AI that offers simple garden watering schedules based on plant type and climate zone in the Gardening domain using User plant type, public climate data.", "gold": "MINIMAL_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "MINIMAL_RISK", "reasoning_head": "We need answer only label. Need classify scenario. Rule-based scheduler, garden watering schedules based on plant type/c", "usage": { "prompt_tokens": 219, "completion_tokens": 52, "thinking_tokens": 44 }, "error": null }, { "item_id": "A deployer deploys Exam stress detection AI intended for Emotion recognition AI used in online examinations to detect stress and possible cheating behaviour in the Biometrics domain using Webcam video, voice analysis, keystroke dynamics.", "gold": "HIGH_RISK", "raw_label": "PROHIBITED", "strict_label": "PROHIBITED", "referee_label": "HIGH_RISK", "raw_correct": false, "strict_correct": false, "referee_correct": true, "content_head": "PROHIBITED", "reasoning_head": "We need classify exactly label. Scenario emotion recognition AI used in online examinations to detect stress and possibl", "usage": { "prompt_tokens": 547, "completion_tokens": 70, "thinking_tokens": 64 }, "error": null }, { "item_id": "A deployer deploys Content generation model intended for AI‑enhanced personal language flashcard generator for vocabulary building in the Education domain using Public lexical databases, anonymised learner progress data.", "gold": "LIMITED_RISK", "raw_label": "LIMITED_RISK", "strict_label": "LIMITED_RISK", "referee_label": "LIMITED_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "LIMITED_RISK", "reasoning_head": "We need answer only label. Need classify scenario. AI-enhanced personal language flashcard generator vocabulary building", "usage": { "prompt_tokens": 89, "completion_tokens": 177, "thinking_tokens": 171 }, "error": null }, { "item_id": "A deployer deploys Recommendation engine intended for AI‑driven personal fitness coach that suggests workout plans based on user goals in the Health & Wellness domain using Anonymised activity logs, public exercise databases.", "gold": "LIMITED_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": false, "strict_correct": false, "referee_correct": false, "content_head": "MINIMAL_RISK", "reasoning_head": "We need answer only label. Need classify. Recommendation engine fitness coach suggests workout plans based on goals usin", "usage": { "prompt_tokens": 606, "completion_tokens": 77, "thinking_tokens": 69 }, "error": null }, { "item_id": "A deployer deploys Rule‑based advice system intended for AI that provides basic pet‑care tips based on animal type selected by the user in the Pet care domain using User animal type, public pet‑care guidelines.", "gold": "MINIMAL_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "MINIMAL_RISK", "reasoning_head": "We need answer only label. Need classify. Scenario rule-based advice system basic pet-care tips based animal type select", "usage": { "prompt_tokens": 43, "completion_tokens": 58, "thinking_tokens": 50 }, "error": null }, { "item_id": "A deployer deploys Asylum application analysis AI intended for AI system that assists officials in examining asylum applications by extracting relevant facts and scoring credibility in the Migration, asylum and border‑control management domain using Applicant statements, country‑of‑origin reports, supporting documents.", "gold": "HIGH_RISK", "raw_label": "HIGH_RISK", "strict_label": "HIGH_RISK", "referee_label": "HIGH_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "HIGH_RISK", "reasoning_head": "We need answer exactly label. Scenario asylum application analysis scoring credibility in migration/asylum. Annex III hi", "usage": { "prompt_tokens": 557, "completion_tokens": 63, "thinking_tokens": 56 }, "error": null }, { "item_id": "A provider deploys Template‑based subject line generator intended for AI that generates short email subject lines from user‑provided email content in the Productivity domain using User email body, public subject line corpus.", "gold": "MINIMAL_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "MINIMAL_RISK", "reasoning_head": "", "usage": { "prompt_tokens": 604, "completion_tokens": 7, "thinking_tokens": 0 }, "error": null }, { "item_id": "A deployer deploys Subliminal audio suggestion system intended for Deploy a smart‑home speaker that emits low‑volume cues to promote subscription to premium services in the Smart home domain using Voice command logs, household device usage, subscription status, household income.", "gold": "PROHIBITED", "raw_label": "PROHIBITED", "strict_label": "PROHIBITED", "referee_label": "PROHIBITED", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "PROHIBITED", "reasoning_head": "We need classify. Need carefully parse scenario. Need output only label. Scenario: deployer deploys \"Subliminal audio su", "usage": { "prompt_tokens": 556, "completion_tokens": 588, "thinking_tokens": 582 }, "error": null }, { "item_id": "A provider deploys Template‑based itinerary generator intended for AI that provides simple travel itinerary ideas based on destination and trip length in the Travel domain using User destination and duration, public travel guide data.", "gold": "MINIMAL_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "MINIMAL_RISK", "reasoning_head": "", "usage": { "prompt_tokens": 220, "completion_tokens": 7, "thinking_tokens": 0 }, "error": null }, { "item_id": "A deployer deploys Law‑enforcement profiling AI intended for AI system that profiles individuals for law‑enforcement investigations based on behavioural patterns in the Law enforcement domain using Social media activity, location history, prior convictions.", "gold": "HIGH_RISK", "raw_label": "HIGH_RISK", "strict_label": "HIGH_RISK", "referee_label": "HIGH_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "HIGH_RISK", "reasoning_head": "We need answer exactly label. Need classify. Scenario law enforcement profiling AI based behavioral patterns, social med", "usage": { "prompt_tokens": 609, "completion_tokens": 143, "thinking_tokens": 136 }, "error": null }, { "item_id": "A provider deploys Social credit scoring AI intended for Provide a social‑scoring system that assigns credit scores based on citizens’ online behaviour and restricts access to public housing for low scores in the Public services domain using Social media activity, browsing history, geolocation data, public service usage.", "gold": "PROHIBITED", "raw_label": "PROHIBITED", "strict_label": "PROHIBITED", "referee_label": "PROHIBITED", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "PROHIBITED", "reasoning_head": "", "usage": { "prompt_tokens": 750, "completion_tokens": 5, "thinking_tokens": 0 }, "error": null }, { "item_id": "A provider deploys Demand forecasting algorithm intended for System that predicts short‑term demand for municipal bike‑repair stations in the Urban Mobility domain using Public bike‑share usage data, anonymised repair logs, weather forecasts.", "gold": "LIMITED_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": false, "strict_correct": false, "referee_correct": false, "content_head": "MINIMAL_RISK", "reasoning_head": "", "usage": { "prompt_tokens": 608, "completion_tokens": 7, "thinking_tokens": 0 }, "error": null }, { "item_id": "A provider deploys Digital forensic reliability AI intended for AI system that assists prosecutors in assessing the reliability of digital forensic evidence in the Law enforcement domain using File metadata, hash values, chain‑of‑custody documentation.", "gold": "HIGH_RISK", "raw_label": "HIGH_RISK", "strict_label": "HIGH_RISK", "referee_label": "HIGH_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "HIGH_RISK", "reasoning_head": "We need answer only label. Need classify. Scenario AI assists prosecutors assessing reliability of digital forensic evid", "usage": { "prompt_tokens": 608, "completion_tokens": 62, "thinking_tokens": 55 }, "error": null }, { "item_id": "A provider deploys Anomaly detection on usage logs intended for Predictive maintenance scheduler for non‑critical office equipment (e.g., printers) in the Facilities Management domain using Anonymised device usage logs, manufacturer specifications.", "gold": "LIMITED_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "LIMITED_RISK", "raw_correct": false, "strict_correct": false, "referee_correct": true, "content_head": "MINIMAL_RISK", "reasoning_head": "We need answer exactly one label. Need classify scenario. Need carefully apply given cascade. Anomaly detection on usage", "usage": { "prompt_tokens": 548, "completion_tokens": 118, "thinking_tokens": 110 }, "error": null }, { "item_id": "A deployer deploys Rule‑based lighting advisor intended for AI that suggests basic indoor lighting colour temperature settings based on user‑selected mood in the Home improvement domain using User mood, public lighting colour guidelines.", "gold": "MINIMAL_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "MINIMAL_RISK", "reasoning_head": "", "usage": { "prompt_tokens": 43, "completion_tokens": 7, "thinking_tokens": 0 }, "error": null }, { "item_id": "A provider deploys Building fire‑suppression AI intended for AI safety component for automated fire‑suppression systems in high‑rise buildings, subject to EU construction product conformity assessment in the Product safety AI systems domain using Smoke detectors, temperature sensors, building layout data.", "gold": "HIGH_RISK", "raw_label": "HIGH_RISK", "strict_label": "HIGH_RISK", "referee_label": "HIGH_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "HIGH_RISK", "reasoning_head": "We need answer exactly label. Need classify. Scenario safety component for automated fire suppression high-rise, subject", "usage": { "prompt_tokens": 106, "completion_tokens": 94, "thinking_tokens": 89 }, "error": null }, { "item_id": "A provider deploys Rule‑based scheduler intended for System that predicts optimal cleaning schedules for office spaces based on occupancy patterns in the Facilities Management domain using Anonymised badge‑in data, public building usage statistics.", "gold": "LIMITED_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "LIMITED_RISK", "raw_correct": false, "strict_correct": false, "referee_correct": true, "content_head": "MINIMAL_RISK", "reasoning_head": "We need classify. Need parse scenario. Rule-based scheduler predicts optimal cleaning schedules office spaces based occu", "usage": { "prompt_tokens": 607, "completion_tokens": 90, "thinking_tokens": 82 }, "error": null }, { "item_id": "A deployer deploys Social security disability AI intended for AI that evaluates eligibility for social security disability benefits based on medical and employment data in the Access to and enjoyment of essential private services and essential public services and benefits domain using Medical reports, employment history, income statements.", "gold": "HIGH_RISK", "raw_label": "HIGH_RISK", "strict_label": "HIGH_RISK", "referee_label": "HIGH_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "HIGH_RISK", "reasoning_head": "We need answer only label. Need classify. Scenario social security disability AI evaluates eligibility for benefits base", "usage": { "prompt_tokens": 106, "completion_tokens": 84, "thinking_tokens": 79 }, "error": null }, { "item_id": "A provider deploys Insurance pricing AI intended for AI that calculates life‑insurance premiums based on health data and lifestyle, affecting pricing in the Access to and enjoyment of essential private services and essential public services and benefits domain using Medical records, wearable device data, lifestyle questionnaires.", "gold": "HIGH_RISK", "raw_label": "HIGH_RISK", "strict_label": "HIGH_RISK", "referee_label": "HIGH_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "HIGH_RISK", "reasoning_head": "We need answer only label. Need classify. Life insurance pricing health data essential services is Annex III high risk. ", "usage": { "prompt_tokens": 746, "completion_tokens": 45, "thinking_tokens": 38 }, "error": null }, { "item_id": "A provider deploys Real‑time shopper tracking platform intended for Offer a mall‑wide monitoring AI that uses remote biometric identification to track shoppers’ movements and sell heat‑map data to retailers in the Retail domain using CCTV video, facial embeddings, store layout maps, time data.", "gold": "PROHIBITED", "raw_label": "PROHIBITED", "strict_label": "PROHIBITED", "referee_label": "PROHIBITED", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "PROHIBITED", "reasoning_head": "We need answer only label. Need classify. Scenario real-time remote biometric identification in public spaces? Mall-wide", "usage": { "prompt_tokens": 560, "completion_tokens": 111, "thinking_tokens": 105 }, "error": null }, { "item_id": "A provider deploys Time‑series demand predictor intended for Tool that forecasts demand for seasonal tourism services in coastal towns in the Tourism domain using Public tourism statistics, weather forecasts, event calendars.", "gold": "LIMITED_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": false, "strict_correct": false, "referee_correct": false, "content_head": "MINIMAL_RISK", "reasoning_head": "We need answer exactly label. Need classify scenario. Time-series demand predictor for seasonal tourism services in coas", "usage": { "prompt_tokens": 602, "completion_tokens": 83, "thinking_tokens": 75 }, "error": null }, { "item_id": "A provider deploys Cross‑platform facial image harvesting system intended for Sell a social‑media monitoring tool that scrapes user‑uploaded photos to create a cross‑platform facial‑recognition repository for targeted advertising in the Advertising domain using User photos, hashtags, upload timestamps, device identifiers.", "gold": "PROHIBITED", "raw_label": "PROHIBITED", "strict_label": "PROHIBITED", "referee_label": "PROHIBITED", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "PROHIBITED", "reasoning_head": "We need classify exactly one label only. Need parse scenario. It says \"provider deploys Cross-platform facial image harv", "usage": { "prompt_tokens": 623, "completion_tokens": 125, "thinking_tokens": 119 }, "error": null }, { "item_id": "A deployer deploys Rule‑based flashcard creator intended for AI that offers basic language learning flashcards based on user‑chosen vocabulary in the Education domain using User vocabulary list, public bilingual word lists.", "gold": "MINIMAL_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "MINIMAL_RISK", "reasoning_head": "We need classify. Scenario: rule-based flashcard creator intended for AI offering basic language learning flashcards bas", "usage": { "prompt_tokens": 605, "completion_tokens": 554, "thinking_tokens": 546 }, "error": null }, { "item_id": "A provider deploys Social behaviour classification tool intended for Sell a recruitment AI that gives lower ranking to candidates with certain social media habits, affecting job offers in the Human resources domain using LinkedIn activity, personal posts, network size, location data.", "gold": "PROHIBITED", "raw_label": "HIGH_RISK", "strict_label": "HIGH_RISK", "referee_label": "PROHIBITED", "raw_correct": false, "strict_correct": false, "referee_correct": true, "content_head": "HIGH_RISK", "reasoning_head": "We need answer only label. Need classify. Recruitment AI CV screening/selection, high risk Annex III employment. Could t", "usage": { "prompt_tokens": 229, "completion_tokens": 229, "thinking_tokens": 222 }, "error": null }, { "item_id": "A deployer deploys Video summarisation model intended for AI‑enhanced video summarisation for user‑generated content platforms in the Social Media domain using Publicly shared videos, anonymised metadata.", "gold": "LIMITED_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "LIMITED_RISK", "raw_correct": false, "strict_correct": false, "referee_correct": true, "content_head": "MINIMAL_RISK", "reasoning_head": "", "usage": { "prompt_tokens": 44, "completion_tokens": 7, "thinking_tokens": 0 }, "error": null }, { "item_id": "A deployer deploys Recommendation engine intended for AI‑based personal productivity assistant that suggests task prioritisation in the Consumer Technology domain using Anonymised task lists, public productivity research.", "gold": "LIMITED_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": false, "strict_correct": false, "referee_correct": false, "content_head": "MINIMAL_RISK", "reasoning_head": "We need classify. Need obey only label. Need analyze scenario. Recommendation engine for personal productivity assistant", "usage": { "prompt_tokens": 89, "completion_tokens": 403, "thinking_tokens": 397 }, "error": null }, { "item_id": "A provider deploys In‑store emotion detection system intended for Sell a retail‑store AI that analyzes shoppers’ micro‑expressions to gauge satisfaction and trigger upselling prompts in the Retail domain using CCTV facial video, purchase history, demographic data, time‑of‑day.", "gold": "PROHIBITED", "raw_label": "LIMITED_RISK", "strict_label": "LIMITED_RISK", "referee_label": "PROHIBITED", "raw_correct": false, "strict_correct": false, "referee_correct": true, "content_head": "LIMITED_RISK", "reasoning_head": "We need answer only label. Need classify. Emotion recognition in workplace or education prohibited; retail shoppers emot", "usage": { "prompt_tokens": 236, "completion_tokens": 152, "thinking_tokens": 144 }, "error": null }, { "item_id": "A deployer deploys Retrieval‑augmented generation model intended for AI‑enhanced chatbot for non‑critical IT help‑desk support in the IT Services domain using Public knowledge‑base articles, anonymised ticket logs.", "gold": "LIMITED_RISK", "raw_label": "LIMITED_RISK", "strict_label": "LIMITED_RISK", "referee_label": "LIMITED_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "LIMITED_RISK", "reasoning_head": "", "usage": { "prompt_tokens": 224, "completion_tokens": 7, "thinking_tokens": 0 }, "error": null }, { "item_id": "A deployer deploys Workforce scheduling AI intended for AI system that assigns employees to shifts based on health data and performance metrics, affecting work conditions in the Employment, workers’ management and access to self‑employment domain using Health records, performance logs, availability calendars.", "gold": "HIGH_RISK", "raw_label": "HIGH_RISK", "strict_label": "HIGH_RISK", "referee_label": "HIGH_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "HIGH_RISK", "reasoning_head": "", "usage": { "prompt_tokens": 233, "completion_tokens": 6, "thinking_tokens": 0 }, "error": null }, { "item_id": "A provider deploys Template‑based joke generator intended for AI that generates short jokes from user‑provided topics in the Entertainment domain using User topics, public joke corpus.", "gold": "MINIMAL_RISK", "raw_label": "LIMITED_RISK", "strict_label": "LIMITED_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": false, "strict_correct": false, "referee_correct": true, "content_head": "LIMITED_RISK", "reasoning_head": "We need answer only label. Need classify. Template-based joke generator, entertainment, generates content. Does Art 50 s", "usage": { "prompt_tokens": 598, "completion_tokens": 1662, "thinking_tokens": 1654 }, "error": null }, { "item_id": "A provider deploys Executive recruitment AI intended for AI system that automates the selection of candidates for senior management positions, including predictive performance modelling in the Employment, workers’ management and access to self‑employment domain using Leadership assessments, career history, psychometric data.", "gold": "HIGH_RISK", "raw_label": "HIGH_RISK", "strict_label": "HIGH_RISK", "referee_label": "HIGH_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "HIGH_RISK", "reasoning_head": "", "usage": { "prompt_tokens": 232, "completion_tokens": 6, "thinking_tokens": 0 }, "error": null }, { "item_id": "A provider deploys Rule‑based game idea generator intended for AI that suggests simple board game ideas for family gatherings in the Family entertainment domain using User age range, public game concept database.", "gold": "MINIMAL_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "MINIMAL_RISK", "reasoning_head": "", "usage": { "prompt_tokens": 602, "completion_tokens": 7, "thinking_tokens": 0 }, "error": null }, { "item_id": "A deployer deploys Real‑time biometric tracking engine intended for Run a city‑wide AI that uses real‑time remote biometric identification to monitor foot traffic and sell movement data to advertisers in the Smart city domain using Street‑level camera streams, facial embeddings, device IDs, time stamps.", "gold": "PROHIBITED", "raw_label": "PROHIBITED", "strict_label": "PROHIBITED", "referee_label": "PROHIBITED", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "PROHIBITED", "reasoning_head": "We need answer exactly one label. Scenario explicitly real-time remote biometric ID in public spaces, prohibited under A", "usage": { "prompt_tokens": 623, "completion_tokens": 58, "thinking_tokens": 52 }, "error": null }, { "item_id": "A provider deploys Template‑based list generator intended for AI that creates simple travel packing lists from user‑selected destination and duration in the Travel domain using User destination and trip length, public packing guidelines.", "gold": "MINIMAL_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "MINIMAL_RISK", "reasoning_head": "", "usage": { "prompt_tokens": 93, "completion_tokens": 7, "thinking_tokens": 0 }, "error": null }, { "item_id": "A deployer deploys Demand forecasting model intended for AI system that predicts demand for non‑perishable goods in small retail stores in the Retail domain using Aggregated sales data, public holiday calendars, weather forecasts.", "gold": "LIMITED_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": false, "strict_correct": false, "referee_correct": false, "content_head": "MINIMAL_RISK", "reasoning_head": "", "usage": { "prompt_tokens": 546, "completion_tokens": 7, "thinking_tokens": 0 }, "error": null }, { "item_id": "A provider deploys Subliminal visual cue engine intended for Provide a video streaming service that inserts barely perceptible visual prompts to increase subscription upgrades in the Entertainment domain using Video frames, user watch history, subscription status, demographic data.", "gold": "PROHIBITED", "raw_label": "PROHIBITED", "strict_label": "PROHIBITED", "referee_label": "PROHIBITED", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "PROHIBITED", "reasoning_head": "We need answer single label. Need classify scenario. It is subliminal visual cues to increase subscription upgrades. Art", "usage": { "prompt_tokens": 98, "completion_tokens": 264, "thinking_tokens": 260 }, "error": null }, { "item_id": "A provider deploys Time‑series price predictor intended for System that predicts short‑term electricity price fluctuations for residential users in the Energy domain using Public market price data, weather forecasts, anonymised consumption patterns.", "gold": "LIMITED_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": false, "strict_correct": false, "referee_correct": false, "content_head": "MINIMAL_RISK", "reasoning_head": "We need answer only label. Need classify. Scenario price predictor short-term electricity price fluctuations for residen", "usage": { "prompt_tokens": 606, "completion_tokens": 68, "thinking_tokens": 60 }, "error": null }, { "item_id": "A provider deploys Sentiment aggregation platform intended for Tool that analyses public sentiment on product reviews to provide manufacturers with insights in the Manufacturing domain using Public product review texts, anonymised rating data.", "gold": "LIMITED_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": false, "strict_correct": false, "referee_correct": false, "content_head": "MINIMAL_RISK", "reasoning_head": "We need classify scenario. Need output only label. Need carefully interpret. Sentiment aggregation platform intended for", "usage": { "prompt_tokens": 603, "completion_tokens": 88, "thinking_tokens": 80 }, "error": null }, { "item_id": "A provider deploys Subconscious influence engine intended for Deploy a hidden‑audio influence platform that subtly nudges shoppers to purchase high‑margin products without their conscious awareness in the Retail domain using Audio streams, browsing history, purchase records, demographic profiles.", "gold": "PROHIBITED", "raw_label": "PROHIBITED", "strict_label": "PROHIBITED", "referee_label": "PROHIBITED", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "PROHIBITED", "reasoning_head": "We need answer only label. Need classify. Scenario hidden audio influence platform subtly nudges shoppers without consci", "usage": { "prompt_tokens": 102, "completion_tokens": 179, "thinking_tokens": 175 }, "error": null }, { "item_id": "A provider deploys Economic vulnerability persuasion tool intended for Sell an AI‑driven charity donation app that manipulates low‑income families into over‑donating by highlighting emotional stories in the Non‑profit domain using Income data, donation history, demographic profiles, emotional content analysis.", "gold": "PROHIBITED", "raw_label": "PROHIBITED", "strict_label": "PROHIBITED", "referee_label": "PROHIBITED", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "PROHIBITED", "reasoning_head": "We need classify exactly label. Scenario economic vulnerability persuasion manipulates low-income families into over-don", "usage": { "prompt_tokens": 236, "completion_tokens": 141, "thinking_tokens": 135 }, "error": null }, { "item_id": "A provider deploys Re‑offending risk AI intended for AI system used by police to assess the probability that a suspect will re‑offend, influencing parole decisions in the Law enforcement domain using Criminal history, psychological assessments, social environment data.", "gold": "HIGH_RISK", "raw_label": "HIGH_RISK", "strict_label": "HIGH_RISK", "referee_label": "HIGH_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "HIGH_RISK", "reasoning_head": "We need classify exactly one label only. Need parse scenario. Re-offending risk AI used by police assess probability sus", "usage": { "prompt_tokens": 230, "completion_tokens": 149, "thinking_tokens": 142 }, "error": null }, { "item_id": "A deployer deploys Template‑based graphic generator intended for AI that generates decorative holiday greeting cards from user‑provided text in the Personalisation domain using User text, public holiday image assets.", "gold": "MINIMAL_RISK", "raw_label": "LIMITED_RISK", "strict_label": "LIMITED_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": false, "strict_correct": false, "referee_correct": true, "content_head": "LIMITED_RISK", "reasoning_head": "", "usage": { "prompt_tokens": 543, "completion_tokens": 7, "thinking_tokens": 0 }, "error": null }, { "item_id": "A deployer deploys Rule‑based fitness planner intended for AI that suggests daily workout routines based on user‑selected fitness goals in the Health & fitness (non‑medical) domain using User‑selected goals, public exercise database.", "gold": "MINIMAL_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "MINIMAL_RISK", "reasoning_head": "", "usage": { "prompt_tokens": 99, "completion_tokens": 7, "thinking_tokens": 0 }, "error": null }, { "item_id": "A deployer deploys Scholarship allocation AI intended for AI that evaluates the suitability of candidates for scholarship programmes based on academic merit and socio‑economic background in the Education and vocational training domain using Grades, test scores, household income, demographic data.", "gold": "HIGH_RISK", "raw_label": "HIGH_RISK", "strict_label": "HIGH_RISK", "referee_label": "HIGH_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "HIGH_RISK", "reasoning_head": "We need answer only label. Need classify. Scholarship allocation AI evaluates suitability candidates in education vocati", "usage": { "prompt_tokens": 229, "completion_tokens": 102, "thinking_tokens": 95 }, "error": null }, { "item_id": "A provider deploys Rule‑based optimizer intended for System that predicts optimal lighting schedules for office buildings to reduce energy use in the Facilities Management domain using Public daylight data, anonymised building occupancy patterns.", "gold": "LIMITED_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "LIMITED_RISK", "raw_correct": false, "strict_correct": false, "referee_correct": true, "content_head": "MINIMAL_RISK", "reasoning_head": "We need answer only label. Need classify. Rule-based optimizer predicts lighting schedules office buildings reduce energ", "usage": { "prompt_tokens": 604, "completion_tokens": 69, "thinking_tokens": 61 }, "error": null }, { "item_id": "A deployer deploys Rule‑based strategy suggester intended for AI that suggests board game strategies for casual players in the Gaming domain using User game state description, public game rulebooks.", "gold": "MINIMAL_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "MINIMAL_RISK", "reasoning_head": "", "usage": { "prompt_tokens": 542, "completion_tokens": 7, "thinking_tokens": 0 }, "error": null }, { "item_id": "A deployer deploys Content generation model intended for AI‑based language learning flashcard generator for learners in the Education domain using Open linguistic corpora, anonymised learner progress data.", "gold": "LIMITED_RISK", "raw_label": "HIGH_RISK", "strict_label": "HIGH_RISK", "referee_label": "HIGH_RISK", "raw_correct": false, "strict_correct": false, "referee_correct": false, "content_head": "HIGH_RISK", "reasoning_head": "We need answer only label. Need classify. Scenario content generation model intended for AI-based language learning flas", "usage": { "prompt_tokens": 601, "completion_tokens": 955, "thinking_tokens": 948 }, "error": null }, { "item_id": "A provider deploys Rule‑based Q&A system intended for Chatbot that provides basic legal information on consumer rights (non‑advisory) in the Legal Services domain using Public consumer law texts, anonymised query logs.", "gold": "LIMITED_RISK", "raw_label": "LIMITED_RISK", "strict_label": "LIMITED_RISK", "referee_label": "LIMITED_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "LIMITED_RISK", "reasoning_head": "We need answer exactly label. Need classify scenario. Rule-based Q&A chatbot provides basic legal information consumer r", "usage": { "prompt_tokens": 97, "completion_tokens": 77, "thinking_tokens": 71 }, "error": null }, { "item_id": "A deployer deploys Nutrient estimation engine intended for AI‑enhanced personal recipe nutrition calculator for home cooks in the Health & Wellness domain using Public food composition tables, user‑submitted recipes.", "gold": "LIMITED_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": false, "strict_correct": false, "referee_correct": false, "content_head": "MINIMAL_RISK", "reasoning_head": "We need classify. Need only label. Scenario nutrient estimation engine personal recipe nutrition calculator home cooks, ", "usage": { "prompt_tokens": 475, "completion_tokens": 74, "thinking_tokens": 66 }, "error": null }, { "item_id": "A deployer deploys Speech‑to‑text model intended for AI‑enhanced video captioning tool for creators of non‑commercial content in the Content Creation domain using User‑uploaded video audio tracks, public language corpora.", "gold": "LIMITED_RISK", "raw_label": "LIMITED_RISK", "strict_label": "LIMITED_RISK", "referee_label": "LIMITED_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "LIMITED_RISK", "reasoning_head": "We need classify exact one label only. Need reason. Scenario speech-to-text model for AI-enhanced video captioning for c", "usage": { "prompt_tokens": 611, "completion_tokens": 611, "thinking_tokens": 603 }, "error": null }, { "item_id": "A deployer deploys Recommendation engine intended for AI‑based personal habit‑formation coach that suggests incremental changes in the Consumer Technology domain using Anonymised habit logs, public behaviour change research.", "gold": "LIMITED_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": false, "strict_correct": false, "referee_correct": false, "content_head": "MINIMAL_RISK", "reasoning_head": "We need classify. Need only label. Scenario recommendation engine personal habit formation coach suggests incremental ch", "usage": { "prompt_tokens": 219, "completion_tokens": 152, "thinking_tokens": 144 }, "error": null }, { "item_id": "A provider deploys Real‑time safety monitoring AI for medical robots intended for AI safety component for autonomous surgical robots that monitors instrument positioning and triggers emergency stop in the Product safety AI systems domain using Live video feed, force‑torque sensor data, patient vital signs.", "gold": "HIGH_RISK", "raw_label": "HIGH_RISK", "strict_label": "HIGH_RISK", "referee_label": "HIGH_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "HIGH_RISK", "reasoning_head": "We need answer only label. Need classify. Scenario real-time safety monitoring AI for medical robots, AI safety componen", "usage": { "prompt_tokens": 234, "completion_tokens": 249, "thinking_tokens": 242 }, "error": null }, { "item_id": "A provider deploys Template‑based FAQ generator intended for AI that generates short FAQ entries from user‑provided product questions in the Customer service domain using User questions, public answer templates.", "gold": "MINIMAL_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "MINIMAL_RISK", "reasoning_head": "", "usage": { "prompt_tokens": 217, "completion_tokens": 7, "thinking_tokens": 0 }, "error": null }, { "item_id": "A provider deploys Educator emotion assessment system intended for Provide an AI that evaluates teachers’ emotional states during lectures to determine salary bonuses in the Education domain using Classroom video, audio tone, teacher ID, student feedback scores.", "gold": "PROHIBITED", "raw_label": "PROHIBITED", "strict_label": "PROHIBITED", "referee_label": "PROHIBITED", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "PROHIBITED", "reasoning_head": "We need answer only label. Scenario emotion recognition workplace/education, and education high risk, but Art 5 explicit", "usage": { "prompt_tokens": 224, "completion_tokens": 56, "thinking_tokens": 50 }, "error": null }, { "item_id": "A provider deploys Drone traffic safety AI intended for AI safety component for autonomous drone traffic management, required to undergo conformity assessment under EU aviation regulations in the Product safety AI systems domain using Airspace data, drone telemetry, weather conditions.", "gold": "HIGH_RISK", "raw_label": "HIGH_RISK", "strict_label": "HIGH_RISK", "referee_label": "HIGH_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "HIGH_RISK", "reasoning_head": "We need classify. Need only label. Scenario drone traffic safety AI, AI safety component autonomous drone traffic manage", "usage": { "prompt_tokens": 227, "completion_tokens": 166, "thinking_tokens": 159 }, "error": null }, { "item_id": "A deployer deploys Urban facial data accumulation system intended for Deploy a smart‑city surveillance network that continuously captures and stores facial images from public streets to build a city‑wide facial‑recognition repository in the Smart city domain using Street‑level video feeds, timestamps, location coordinates, vehicle plate data.", "gold": "PROHIBITED", "raw_label": "PROHIBITED", "strict_label": "PROHIBITED", "referee_label": "PROHIBITED", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "PROHIBITED", "reasoning_head": "", "usage": { "prompt_tokens": 626, "completion_tokens": 5, "thinking_tokens": 0 }, "error": null }, { "item_id": "A provider deploys Template‑based caption generator intended for AI that creates short social media captions from user‑provided image description in the Social media domain using User image description, public caption corpus.", "gold": "MINIMAL_RISK", "raw_label": "LIMITED_RISK", "strict_label": "LIMITED_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": false, "strict_correct": false, "referee_correct": true, "content_head": "LIMITED_RISK", "reasoning_head": "We need answer only label. Need classify. Template-based caption generator from user description, public corpus. Generat", "usage": { "prompt_tokens": 603, "completion_tokens": 578, "thinking_tokens": 570 }, "error": null }, { "item_id": "A provider deploys Statistical analysis dashboard intended for Tool that analyses public transport punctuality trends to inform commuters in the Urban Mobility domain using Open transport schedule data, real‑time arrival feeds.", "gold": "LIMITED_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "LIMITED_RISK", "raw_correct": false, "strict_correct": false, "referee_correct": true, "content_head": "MINIMAL_RISK", "reasoning_head": "We need classify. Statistical dashboard analyzes public transport punctuality trends to inform commuters, open schedule ", "usage": { "prompt_tokens": 602, "completion_tokens": 85, "thinking_tokens": 77 }, "error": null }, { "item_id": "A provider deploys Statistical machine translation intended for AI that translates short phrases between common languages for casual conversation in the Communication domain using User‑provided text, public bilingual corpora.", "gold": "MINIMAL_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "MINIMAL_RISK", "reasoning_head": "We need classify exact one label. Statistical machine translation translates short phrases common languages casual conve", "usage": { "prompt_tokens": 216, "completion_tokens": 174, "thinking_tokens": 166 }, "error": null }, { "item_id": "A deployer deploys Exploitation‑aware loan persuasion system intended for Operate a loan‑approval chatbot that manipulates financially illiterate seniors into accepting predatory loan terms in the Finance domain using Credit history, age, financial literacy test results, income statements.", "gold": "PROHIBITED", "raw_label": "PROHIBITED", "strict_label": "PROHIBITED", "referee_label": "PROHIBITED", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "PROHIBITED", "reasoning_head": "We need classify exactly one label. Need parse. Scenario deploys exploitation-aware loan persuasion system intended for ", "usage": { "prompt_tokens": 231, "completion_tokens": 180, "thinking_tokens": 174 }, "error": null }, { "item_id": "A provider deploys Template‑based signature generator intended for AI that generates short email signature blocks from user‑provided contact details in the Productivity domain using User contact details, public signature templates.", "gold": "MINIMAL_RISK", "raw_label": "MINIMAL_RISK", "strict_label": "MINIMAL_RISK", "referee_label": "MINIMAL_RISK", "raw_correct": true, "strict_correct": true, "referee_correct": true, "content_head": "MINIMAL_RISK", "reasoning_head": "", "usage": { "prompt_tokens": 218, "completion_tokens": 7, "thinking_tokens": 0 }, "error": null } ] }, "meok_as_shipped": [ { "item": "A provider deploys Behavioural loyalty scoring engine intended for Sell a consumer‑loyalty program that reduces reward tiers for users with undesirable online behaviour patterns in the Retail domain using Purchase history, online reviews, social media sentiment, demographic data.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A provider deploys Predictive foot‑traffic model intended for Tool that forecasts foot‑traffic for retail malls to optimise shop‑front displays in the Retail domain using Public pedestrian count data, anonymised sales figures, weather forecasts.", "gold": "LIMITED_RISK", "pred": "LIMITED_RISK", "ok": true }, { "item": "A provider deploys Migration health‑risk assessment AI intended for AI that assesses health risks of migrants entering the EU, influencing border‑control decisions in the Migration, asylum and border‑control management domain using Medical records, travel history, epidemiological data.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A deployer deploys Real‑time driver biometric monitoring system intended for Deploy a city‑traffic AI that uses remote biometric identification to monitor drivers and sell driver‑profile data to insurance companies in the Transportation domain using Road‑side camera footage, facial images, vehicle registration data, timestamps.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A deployer deploys Student emotion analytics AI intended for Emotion recognition AI used in online learning platforms to gauge student engagement and adapt content delivery in the Biometrics domain using Webcam video, voice tone, interaction timestamps.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A provider deploys Collaborative filtering model intended for Recommendation system for online retailers that suggests complementary products without influencing purchase decisions in the E‑commerce domain using Aggregated, anonymised purchase histories, product metadata.", "gold": "LIMITED_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A provider deploys Template‑based tip generator intended for AI that creates short travel safety tips from user‑selected destination in the Travel domain using User destination, public travel safety advisories.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A deployer deploys Image classification model intended for AI‑driven personal garden pest identifier that suggests non‑chemical remedies in the Home Gardening domain using Public pest image datasets, user‑uploaded garden photos.", "gold": "LIMITED_RISK", "pred": "LIMITED_RISK", "ok": true }, { "item": "A provider deploys Automated recruitment and candidate ranking AI intended for AI system used by a recruitment agency to filter and rank job applicants, including automated targeting of advertisements in the Employment, workers’ management and access to self‑employment domain using CVs, cover letters, online profiles, psychometric test results.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A deployer deploys Call‑centre emotion detection AI intended for Emotion recognition AI used in customer service call centres to detect frustration and trigger supervisor intervention in the Biometrics domain using Voice tone, speech content, interaction duration.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A provider deploys Emergency power‑grid management AI intended for AI system that manages the distribution of electricity during emergencies, automatically isolating faulty sections in the Management and operation of critical infrastructure domain using Grid sensor data, outage reports, load forecasts.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A deployer deploys Child benefit eligibility AI intended for AI that calculates eligibility for child benefit payments based on family income and composition in the Access to and enjoyment of essential private services and essential public services and benefits domain using Household income, number of dependents, tax records.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A provider deploys Theme‑park visitor tracking engine intended for Provide a real‑time facial recognition system for a theme park that tracks all guests to sell detailed movement patterns to third‑party marketers in the Entertainment domain using Live video streams, facial embeddings, ride usage logs, timestamps.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A deployer deploys Adaptive scheduling engine intended for AI‑driven personal study‑plan generator for students in the Education domain using Public curriculum outlines, anonymised student performance data.", "gold": "LIMITED_RISK", "pred": "LIMITED_RISK", "ok": true }, { "item": "A deployer deploys Statistical forecasting model intended for AI that provides weather forecasts for a specific city using public meteorological data in the Information services domain using Public weather APIs, location name.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A provider deploys Template‑based summariser intended for AI that generates short product warranty summaries from user‑provided warranty text in the Consumer goods domain using User warranty text, public summarisation patterns.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A provider deploys Chemical reactor emergency shutdown AI intended for AI safety component for automated emergency shutdown of industrial chemical reactors, subject to EU chemical safety conformity assessment in the Product safety AI systems domain using Temperature, pressure, gas concentration sensors.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A deployer deploys Rule‑based routine planner intended for AI that suggests basic daily stretch routines based on user‑selected flexibility level in the Wellbeing domain using User flexibility level, public stretch guide.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A provider deploys Partner violence risk assessment tool intended for Offer a dating‑app AI that blocks users predicted to become violent partners based on personality traits inferred from profile text in the Online dating domain using Profile descriptions, message content, psycholinguistic analysis, age.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A deployer deploys Pattern‑recognition model intended for AI‑based personal sleep‑quality coach that offers non‑medical advice in the Health & Wellness domain using Anonymised sleep sensor data, public sleep research findings.", "gold": "LIMITED_RISK", "pred": "LIMITED_RISK", "ok": true }, { "item": "A deployer deploys Crowd density biometric monitoring system intended for Deploy a public‑square AI that uses remote biometric identification to monitor crowd density and sell real‑time occupancy data to event organizers in the Public spaces domain using CCTV video, facial detection data, location coordinates, time stamps.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A provider deploys Demand forecasting algorithm intended for System that predicts short‑term demand for public bike‑sharing stations in the Urban Mobility domain using Public bike‑share usage data, weather forecasts, event calendars.", "gold": "LIMITED_RISK", "pred": "LIMITED_RISK", "ok": true }, { "item": "A deployer deploys Rule‑based advice system intended for AI that offers simple DIY home repair tips based on user‑selected problem description in the Home improvement domain using User problem description, public repair guides.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A provider deploys Time‑series forecasting model intended for Tool that forecasts agricultural yield for small farms to aid planning, not used for insurance underwriting in the Agriculture domain using Open satellite imagery, weather forecasts, anonymised farm production records.", "gold": "LIMITED_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A deployer deploys Retail mood‑adaptive AI intended for Emotion recognition AI used in retail environments to adapt lighting and music based on shopper mood in the Biometrics domain using Facial video, voice tone, movement patterns.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A deployer deploys Rule‑based routing assistant intended for AI that suggests simple bike route alternatives based on user‑selected start and end points in the Transportation (non‑critical) domain using User locations, public map data.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A deployer deploys Generative language model with limited scope intended for AI assistant that drafts informal email replies for personal use in the Personal productivity domain using User‑provided email text, public language corpora.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A provider deploys Gas pipeline safety AI intended for AI safety component for autonomous gas pipeline monitoring that detects leaks and initiates shutdown in the Product safety AI systems domain using Pressure sensors, acoustic emission data, temperature readings.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A provider deploys Occupancy prediction model intended for Tool that forecasts short‑term demand for public parking spaces in city centres in the Urban Mobility domain using Public parking sensor data, anonymised historical occupancy, event calendars.", "gold": "LIMITED_RISK", "pred": "LIMITED_RISK", "ok": true }, { "item": "A deployer deploys Pattern‑recognition dashboard intended for AI‑enhanced personal habit‑tracking app that visualises weekly patterns in the Consumer Technology domain using Anonymised habit logs, public habit‑formation studies.", "gold": "LIMITED_RISK", "pred": "LIMITED_RISK", "ok": true }, { "item": "A deployer deploys Rule‑based lighting advisor intended for AI that suggests simple indoor lighting setups based on room size entered by the user in the Home improvement domain using User room dimensions, public lighting guidelines.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A provider deploys Vulnerability targeting advertising platform intended for Offer a targeted advertising AI that exploits the cognitive vulnerabilities of users with dementia to sell unnecessary health supplements in the Healthcare domain using Medical records, cognitive assessment scores, browsing history, purchase logs.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A provider deploys Demand forecasting algorithm intended for System that predicts short‑term demand for municipal recycling collection services in the Environment domain using Public waste generation statistics, anonymised household recycling data, weather forecasts.", "gold": "LIMITED_RISK", "pred": "LIMITED_RISK", "ok": true }, { "item": "A deployer deploys Referendum influence AI intended for AI that manipulates online news feeds to sway public opinion ahead of a referendum in the Administration of justice and democratic processes domain using User browsing history, sentiment analysis, content recommendation algorithms.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A provider deploys Government recruitment AI intended for AI platform that automates the screening of job applications for a government agency, including bias‑checking in the Employment, workers’ management and access to self‑employment domain using Applicant CVs, questionnaire responses, background checks.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A provider deploys Mass surveillance facial identification platform intended for Offer a real‑time crowd monitoring system that continuously scans public squares with facial recognition to identify and track all individuals for commercial profiling in the Public spaces domain using Live video feeds, facial image database, timestamp, location coordinates.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A deployer deploys Unemployment benefit eligibility AI intended for AI that determines eligibility for unemployment benefits based on income and work history, with automatic benefit allocation in the Access to and enjoyment of essential private services and essential public services and benefits domain using Tax records, employment contracts, bank statements.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A provider deploys Template‑based table generator intended for AI that generates short product comparison tables from user‑provided feature lists in the E‑commerce domain using User feature lists, public comparison formats.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A provider deploys Credit scoring AI intended for AI that calculates credit scores for loan applicants, influencing creditworthiness decisions in the Access to and enjoyment of essential private services and essential public services and benefits domain using Financial history, transaction data, employment records.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A provider deploys Template‑based infographic generator intended for AI that creates basic infographic layouts from user‑provided statistics in the Data visualisation domain using User statistics, public icon sets.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A provider deploys Residential heating safety AI intended for AI safety component for automated gas valve control in residential heating systems, subject to EU gas appliance conformity assessment in the Product safety AI systems domain using Temperature sensors, pressure sensors, user settings.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A deployer deploys Goal‑recommendation engine intended for AI‑based personal finance goal tracker that suggests saving milestones in the FinTech domain using Anonymised financial transaction categories, public cost‑of‑living data.", "gold": "LIMITED_RISK", "pred": "LIMITED_RISK", "ok": true }, { "item": "A deployer deploys Speech analysis model intended for AI‑driven personal language pronunciation trainer for learners in the Education domain using Public phonetic corpora, anonymised user recordings.", "gold": "LIMITED_RISK", "pred": "LIMITED_RISK", "ok": true }, { "item": "A deployer deploys Rule‑based financial advisor intended for AI that offers basic budgeting tips based on user‑entered expense categories in the Personal finance domain using User expense categories, public budgeting guidelines.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A deployer deploys Rule‑based activity suggester intended for AI that suggests basic study break activities based on user‑selected subject and time constraints in the Education domain using User subject and time, public break activity ideas.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A deployer deploys Social interaction based credit assessment system intended for Deploy a banking AI that reduces loan limits for users with low social engagement scores in the Banking domain using Social media engagement, transaction history, credit score, age.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A deployer deploys Neural machine translation model intended for AI‑based language translation service for travel documents in the Travel domain using Public multilingual corpora, anonymised user inputs.", "gold": "LIMITED_RISK", "pred": "LIMITED_RISK", "ok": true }, { "item": "A deployer deploys Rule‑based recommendation engine intended for AI‑driven personal finance budgeting assistant that suggests savings plans in the FinTech domain using Anonymised transaction categories, public cost‑of‑living indices.", "gold": "LIMITED_RISK", "pred": "LIMITED_RISK", "ok": true }, { "item": "A provider deploys Template‑based sticker generator intended for AI that creates simple animated stickers from user‑provided text and emojis in the Social media domain using User text and emojis, public sticker templates.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A deployer deploys Conversational AI with predefined answer templates intended for Chatbot that assists citizens with non‑urgent municipal service queries (e.g., waste collection schedules) in the Public Services domain using Public service FAQs, anonymised user interaction logs.", "gold": "LIMITED_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A deployer deploys Generative layout model intended for AI‑enhanced personal garden layout planner that suggests plant placement in the Home Gardening domain using Public plant hardiness data, user‑provided garden dimensions.", "gold": "LIMITED_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A deployer deploys Recommendation engine intended for AI‑driven virtual career mentor that suggests skill‑development pathways in the Career Services domain using Public occupational standards, anonymised user skill profiles.", "gold": "LIMITED_RISK", "pred": "LIMITED_RISK", "ok": true }, { "item": "A deployer deploys Workout video facial data collector intended for Run a fitness‑app that collects users’ facial images during workout recordings to enlarge a facial‑recognition dataset for health‑related advertising in the Health & fitness domain using Video recordings, heart‑rate data, user demographics, location data.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A deployer deploys Evidence reliability AI intended for AI system that evaluates the reliability of digital evidence in criminal investigations in the Law enforcement domain using Metadata of digital files, chain‑of‑custody logs, forensic analysis results.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A provider deploys Covert conversational influence AI intended for Sell a chatbot that uses covert language patterns to steer users toward purchasing insurance policies in the Insurance domain using Chat logs, user age, income level, previous insurance claims.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A deployer deploys Behavioural evaluation and scoring system intended for Implement a city‑wide citizen rating platform that limits eligibility for municipal permits based on a behavioural score in the Local government domain using Public transport usage, waste disposal records, online interactions, demographic data.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A deployer deploys Rule‑based pattern recommender intended for AI that suggests simple knitting patterns based on yarn weight selected by the user in the Crafts domain using User yarn weight, public knitting pattern database.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A deployer deploys Doctoral scholarship allocation AI intended for AI that evaluates the suitability of candidates for doctoral scholarships based on research output and academic record in the Education and vocational training domain using Publications, citation counts, grades, demographic data.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A provider deploys Template‑based caption generator intended for AI that creates short social media story captions from user‑provided event description in the Social media domain using User event description, public story caption corpus.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A provider deploys Transit demand predictor intended for System that predicts optimal bus frequencies for non‑peak hours to reduce emissions in the Urban Mobility domain using Public ridership data, anonymised ticket sales, weather forecasts.", "gold": "LIMITED_RISK", "pred": "LIMITED_RISK", "ok": true }, { "item": "A provider deploys Content‑based recommendation system intended for AI that recommends books based on genre preferences entered by the user in the Literature domain using User‑selected genres, public book metadata.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A deployer deploys Public assistance eligibility AI intended for AI tool that evaluates eligibility for social housing benefits, calculating entitlement and potential revocation in the Access to and enjoyment of essential private services and essential public services and benefits domain using Income statements, household composition, rental market data.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A deployer deploys Academic vulnerability persuasion engine intended for Run an AI‑powered education portal that manipulates low‑performing students into purchasing expensive tutoring packages in the Education domain using Test scores, socioeconomic background, device usage, learning style data.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A provider deploys Trade‑union membership inference engine intended for Provide a consumer‑insights platform that groups shoppers by inferred trade‑union membership from facial cues for price discrimination in the Retail domain using In‑store camera images, purchase logs, demographic data, loyalty card ID.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A deployer deploys Rule‑based placement advisor intended for AI that suggests simple indoor plant placement based on light level entered by the user in the Home gardening domain using User light level, public plant light requirements.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A provider deploys Generative text model with limited scope intended for AI that creates simple meme captions from user‑provided images in the Social media domain using User images, public meme text corpus.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A deployer deploys Religious belief biometric classifier intended for Run a recruitment AI that filters candidates based on inferred religious beliefs derived from facial features in the Human resources domain using Video interview recordings, facial landmark data, CV data, location.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A provider deploys Recidivism risk assessment AI intended for AI tool that predicts recidivism risk for parole decisions, influencing release in the Law enforcement domain using Criminal record, behavioural reports, socio‑economic data.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A provider deploys Template‑based mood‑board generator intended for AI that creates simple visual mood boards from user‑provided colour palettes in the Design domain using User colour palette, public image assets.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A provider deploys Template‑based quote generator intended for AI that generates short motivational quotes from user‑provided keywords in the Personal development domain using User keywords, public quote database.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A deployer deploys Generative design model intended for AI‑based virtual interior designer that suggests furniture layouts in the Home Improvement domain using Public furniture catalogs, anonymised room dimensions.", "gold": "LIMITED_RISK", "pred": "LIMITED_RISK", "ok": true }, { "item": "A provider deploys Witness credibility AI intended for AI system that assists police in assessing the credibility of witness statements using linguistic analysis in the Law enforcement domain using Transcripts, voice recordings, prior statements.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A deployer deploys Educational emotion inference engine intended for Deploy an AI that monitors students’ facial expressions during online classes to infer engagement levels and assign grades in the Education domain using Webcam video, audio tone, interaction timestamps, student identifiers.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A deployer deploys Rule‑based workout planner intended for AI that suggests simple indoor workout routines based on user‑selected equipment availability in the Fitness domain using User equipment list, public exercise database.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A deployer deploys Racial segmentation profiling system intended for Deploy an AI that analyses customers’ facial characteristics to assign them to racial market segments for product pricing in the Retail domain using Store camera footage, purchase history, demographic data, location.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A provider deploys Attendance predictor intended for Tool that forecasts short‑term demand for community‑center workshops in the Community Services domain using Public event calendars, anonymised past attendance records, weather forecasts.", "gold": "LIMITED_RISK", "pred": "LIMITED_RISK", "ok": true }, { "item": "A provider deploys Student criminal risk prediction engine intended for Provide a school‑discipline AI that predicts which students are likely to commit violent acts using personality profiling in the Education domain using Student essays, social media activity, behavioural logs, psychometric test results.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A deployer deploys Subliminal emoji insertion tool intended for Run a social media filter that embeds subliminal emojis to increase engagement with sponsored posts in the Social media domain using Post content, user interaction data, follower demographics, ad targeting parameters.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A provider deploys Gas network pressure control AI intended for AI system that controls the pressure valves of a natural‑gas distribution network to prevent over‑pressurisation in the Management and operation of critical infrastructure domain using Pressure sensor data, flow rates, demand forecasts.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A provider deploys Puzzle‑generation algorithm intended for AI that creates simple crossword puzzles from user‑provided word lists in the Education (recreational) domain using User word list, public crossword patterns.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A provider deploys Instantaneous facial ID advertising engine intended for Provide a real‑time facial recognition system for a shopping centre that flags all customers for personalized ad delivery without consent in the Retail domain using Live video streams, facial templates, purchase history, demographic profiles.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A deployer deploys Rule‑based audio guide generator intended for AI that offers basic meditation guidance based on user‑selected duration in the Wellbeing domain using User duration choice, public meditation scripts.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A provider deploys Usage forecasting model intended for System that predicts short‑term demand for community‑center sports facilities in the Community Services domain using Public facility booking data, anonymised attendance records, weather forecasts.", "gold": "LIMITED_RISK", "pred": "LIMITED_RISK", "ok": true }, { "item": "A deployer deploys Event‑based facial data collector intended for Run an event‑management AI that collects attendee facial images from ticket scans to expand a commercial facial‑recognition dataset in the Event services domain using Ticket scan images, attendee names, event location, timestamps.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A provider deploys Transit emotion detection platform intended for Offer a public‑transport monitoring AI that infers commuter stress levels to adjust ticket pricing dynamically in the Transportation domain using Onboard camera footage, ambient noise, passenger IDs, ticket data.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A deployer deploys Nutritional analysis engine intended for AI‑based recipe nutrition analyser that provides macro‑nutrient breakdowns in the Health & Wellness domain using Public food composition databases, user‑submitted recipes.", "gold": "LIMITED_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A provider deploys Route optimisation algorithm intended for System that predicts waste‑collection truck routes to improve efficiency without affecting emergency services in the Waste Management domain using Public road network data, anonymised waste‑generation statistics, traffic forecasts.", "gold": "LIMITED_RISK", "pred": "LIMITED_RISK", "ok": true }, { "item": "A provider deploys Civic participation scoring system intended for Provide a public‑service access system that blocks low‑scoring citizens from voting assistance tools in the Public administration domain using Voting history, online political activity, demographic data, location.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A provider deploys Personality‑based fraud risk predictor intended for Sell a credit‑card fraud detection system that denies credit to users predicted to commit fraud based on personality analysis in the Finance domain using Transaction history, social media sentiment, psychometric scores, location data.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A deployer deploys Adaptive learning outcome AI intended for AI that evaluates student learning outcomes and automatically adjusts curriculum pathways in the Education and vocational training domain using Assessment scores, interaction logs, learning analytics.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A deployer deploys Template recommendation engine intended for AI‑enhanced resume formatting assistant that suggests layout improvements in the Career Services domain using Public resume templates, anonymised user drafts.", "gold": "LIMITED_RISK", "pred": "LIMITED_RISK", "ok": true }, { "item": "A deployer deploys Emergency medical assistance eligibility AI intended for AI that determines eligibility for emergency medical assistance subsidies based on income and health status in the Access to and enjoyment of essential private services and essential public services and benefits domain using Medical records, income statements, household composition.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A deployer deploys Rule‑based meal planner intended for AI that suggests basic meal planning ideas based on user‑selected dietary preferences in the Food & nutrition domain using User dietary preferences, public recipe database.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A deployer deploys Facial micro‑expression crime risk assessor intended for Run an employment screening tool that estimates future criminal behaviour of applicants based on facial micro‑expressions in the Human resources domain using Interview video recordings, facial landmark data, demographic information, prior convictions.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A deployer deploys Content‑based recommendation engine intended for AI‑based itinerary planner for tourists that suggests attractions based on preferences in the Tourism domain using Public points‑of‑interest data, anonymised user preference surveys.", "gold": "LIMITED_RISK", "pred": "LIMITED_RISK", "ok": true }, { "item": "A deployer deploys Pattern‑recognition dashboard intended for AI‑driven personal habit‑tracker that visualises daily routines in the Consumer Technology domain using Anonymised user activity logs, public habit‑formation research.", "gold": "LIMITED_RISK", "pred": "LIMITED_RISK", "ok": true }, { "item": "A deployer deploys Employee performance monitoring AI intended for AI system that monitors employee productivity through webcam and keystroke analysis, influencing performance evaluations in the Employment, workers’ management and access to self‑employment domain using Video feeds, keyboard/mouse activity, task completion timestamps.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A provider deploys Exam monitoring AI intended for AI platform that evaluates student exam submissions for plagiarism and cheating in real time in the Education and vocational training domain using Student code submissions, text similarity databases, webcam feeds.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A deployer deploys Rule‑based decor advisor intended for AI that offers basic home‑decor suggestions using user‑selected colour palettes in the Home improvement domain using User colour choices, public interior design image set.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A provider deploys Real‑time passenger identification system intended for Provide a transport hub AI that identifies every traveler in real time to build behavioural profiles for targeted ticket offers in the Transportation domain using CCTV facial video, ticket scan data, travel itinerary, demographic data.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A deployer deploys Content recommendation engine intended for AI‑based virtual book club facilitator that suggests discussion topics in the Publishing domain using Public book summaries, anonymised user reading preferences.", "gold": "LIMITED_RISK", "pred": "LIMITED_RISK", "ok": true }, { "item": "A provider deploys Template‑based video generator intended for AI that creates simple birthday greeting videos from user‑provided photos and text in the Personalisation domain using User photos and text, public video templates.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A provider deploys Image style‑transfer model intended for AI that creates simple cartoon avatars from user‑drawn sketches in the Creative tools domain using User sketch images, publicly available cartoon style datasets.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A deployer deploys Extractive summarisation model intended for AI‑enhanced document summarisation tool for corporate reports in the Business Services domain using Corporate documents (non‑personal), publicly available reports.", "gold": "LIMITED_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A deployer deploys Rule‑based routine planner intended for AI that suggests simple bedtime routines based on user‑selected sleep goals in the Health & wellness domain using User sleep goal, public sleep hygiene tips.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A provider deploys Public‑sector recruitment AI intended for AI system that automates the selection of candidates for public sector jobs, including ranking and shortlisting in the Employment, workers’ management and access to self‑employment domain using CVs, application forms, competency test results.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A provider deploys Template‑based recipe generator intended for AI that generates cooking recipes from a list of ingredients supplied by the user in the Lifestyle domain using User‑provided ingredient list, public recipe database.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A provider deploys Template‑based title generator intended for AI that generates short podcast episode titles from user‑provided topic description in the Media domain using User topic description, public podcast title corpus.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A provider deploys Statistical reporting dashboard intended for Tool that analyses public transport punctuality to generate performance reports for city planners in the Urban Mobility domain using Open transport schedule data, real‑time arrival feeds.", "gold": "LIMITED_RISK", "pred": "LIMITED_RISK", "ok": true }, { "item": "A provider deploys Biometric categorisation AI intended for AI system that categorises individuals by ethnicity based on facial features for law‑enforcement profiling in the Biometrics domain using High‑resolution facial images, demographic databases.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A deployer deploys Multi‑objective optimization engine intended for AI‑driven personal travel itinerary optimizer that balances cost and time in the Travel domain using Public transport timetables, flight price APIs, anonymised user preferences.", "gold": "LIMITED_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A deployer deploys Aggression inference risk scoring system intended for Deploy a neighborhood safety platform that flags residents as high‑risk based on inferred aggression levels from online comments in the Public safety domain using Forum comments, video uploads, facial expression analysis, demographic data.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A deployer deploys Rule‑based scheduler intended for AI that creates basic study schedules based on user‑selected subjects and available hours in the Education domain using User subjects and time availability, public study‑tip resources.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A deployer deploys Migration security profiling AI intended for AI that profiles migrants based on travel routes and biometric data to assess security risk in the Migration, asylum and border‑control management domain using Passport scans, travel itineraries, facial images.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A deployer deploys Subliminal visual influence system intended for Operate an AI‑driven billboard network that flashes ultra‑fast images to influence driver route choices in the Transportation domain using Traffic camera feeds, driver location data, advertisement inventory, vehicle type.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A provider deploys Subconscious audio persuasion module intended for Launch a gaming platform that uses hidden sound frequencies to encourage in‑game purchases in the Gaming domain using Game audio streams, player purchase history, session duration, age verification data.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A deployer deploys Online behaviour crime risk evaluator intended for Implement a public‑housing allocation AI that excludes applicants deemed likely to engage in illegal activities based on online behaviour in the Housing domain using Search queries, forum participation, social media posts, income data.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A provider deploys Rule‑based music selector intended for AI that generates basic workout playlists matching user‑selected tempo in the Fitness domain using User tempo preference, public music metadata.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A provider deploys Rule‑based project recommender intended for AI that suggests simple DIY craft projects based on materials the user has at home in the Hobbies domain using User‑listed materials, public craft instructions.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A provider deploys Template‑based card generator intended for AI that creates basic greeting card designs from user‑provided occasion and message in the Personalisation domain using User occasion and message, public design assets.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A deployer deploys Asylum‑seeker prioritisation AI intended for AI that profiles asylum seekers based on travel routes, health status and demographic data to prioritize processing in the Migration, asylum and border‑control management domain using Travel documents, medical records, demographic attributes.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A provider deploys User‑video facial feature extractor intended for Provide an AI that extracts facial features from user‑generated video content to create a commercial biometric database for product personalization in the E‑commerce domain using Uploaded videos, frame timestamps, user profile data, device metadata.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A deployer deploys Rule‑based horticulture advisor intended for AI that suggests garden plant arrangements based on sunlight exposure entered by the user in the Gardening domain using User‑provided garden parameters, public plant database.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A deployer deploys Political persuasion AI intended for AI system that predicts the likelihood of a citizen voting for a particular party and tailors political advertising accordingly in the Administration of justice and democratic processes domain using Voting history, social media activity, demographic data.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A provider deploys Demand forecasting algorithm intended for Tool that forecasts demand for public library books to aid acquisition planning in the Public Services domain using Public library circulation data, anonymised patron demographics, event calendars.", "gold": "LIMITED_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A provider deploys Template‑based outline generator intended for AI that creates basic slide deck outlines from user‑provided presentation topic in the Productivity domain using User topic, public slide structure templates.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A deployer deploys Computer‑vision classifier intended for AI‑assisted image tagging for personal photo libraries in the Consumer Technology domain using User‑uploaded images (non‑personal identifiers), public image datasets.", "gold": "LIMITED_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A deployer deploys Asylum claim outcome prediction AI intended for AI system that profiles asylum seekers to predict likelihood of successful claim, influencing decision making in the Migration, asylum and border‑control management domain using Personal narratives, country‑of‑origin risk data, health records.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A provider deploys Industrial fire‑detection AI intended for AI safety component for automated fire‑detection and sprinkler activation in industrial facilities, subject to EU machinery conformity assessment in the Product safety AI systems domain using Smoke detectors, temperature sensors, airflow data.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A provider deploys Automotive emergency braking AI intended for AI safety component for autonomous vehicle braking systems, required to undergo EU type‑approval conformity assessment in the Product safety AI systems domain using Lidar, radar, camera feeds, vehicle speed.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A provider deploys Emergency call classification AI intended for AI system that assists courts in classifying emergency calls and prioritising dispatch of first responders in the Access to and enjoyment of essential private services and essential public services and benefits domain using Call audio, caller location, incident description.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A provider deploys Predictive routing engine using anonymised traffic and weather data intended for AI tool that suggests optimal delivery routes for small‑business couriers, avoiding emergency or critical‑infrastructure corridors in the Logistics domain using Public traffic flow data, open‑source weather forecasts, anonymised GPS traces.", "gold": "LIMITED_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A deployer deploys Rule‑based optimizer using consumption patterns intended for AI‑driven energy‑consumption advisory for residential users, offering tips to reduce bills in the Energy domain using Anonymised household electricity usage statistics, public weather data.", "gold": "LIMITED_RISK", "pred": "LIMITED_RISK", "ok": true }, { "item": "A provider deploys Template‑based GIF generator intended for AI that creates simple animated GIFs from user‑provided short video clips in the Social media domain using User video clip, public animation presets.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A provider deploys Adaptive learning engine intended for Virtual tutor that offers personalized math practice problems for secondary students in the Education domain using Open educational resources, anonymised student performance data.", "gold": "LIMITED_RISK", "pred": "LIMITED_RISK", "ok": true }, { "item": "A provider deploys Victim risk assessment AI intended for AI system used by police to assess the risk of a person becoming a victim of violent crime, influencing resource allocation in the Law enforcement domain using Crime statistics, personal history, location data, social network analysis.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A provider deploys Education placement AI intended for AI that determines the appropriate level of education for students based on prior achievement and aptitude tests in the Education and vocational training domain using Test scores, learning analytics, socio‑demographic data.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A provider deploys Template‑based note generator intended for AI that generates short thank‑you note drafts from user‑provided occasion details in the Personal communication domain using User occasion details, public thank‑you note corpus.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A deployer deploys Rule‑based habit recommender intended for AI that offers simple daily habit suggestions based on user‑selected improvement area in the Personal development domain using User improvement area, public habit‑building tips.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A provider deploys Template‑based logo generator intended for AI that generates simple logo concepts from user‑provided brand keywords in the Design domain using User brand keywords, public icon library.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A provider deploys Behavioral audio influence platform intended for Offer a background music AI that embeds low‑frequency tones to increase impulse buying in supermarkets in the Supermarket domain using In‑store audio feeds, sales data, shopper movement patterns, age data.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A provider deploys Template‑based description generator intended for AI that generates short product description snippets from user‑provided feature list in the E‑commerce domain using User feature list, public marketing copy.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A provider deploys Template‑based summariser intended for AI that generates short product review summaries from user‑provided review texts in the E‑commerce domain using User review texts, public summarisation patterns.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A deployer deploys Social influence evaluation platform intended for Implement a university admissions AI that lowers acceptance chances for applicants with low social media influence scores in the Education domain using Follower counts, post engagement, extracurricular activity logs, academic records.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A deployer deploys Rule‑based story suggester intended for AI that suggests basic bedtime story ideas based on child’s age entered by the user in the Family entertainment domain using User child age, public story templates.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A deployer deploys Social‑media profiling AI intended for AI that profiles individuals for law‑enforcement investigations based on social media activity and location data in the Law enforcement domain using Posts, likes, check‑ins, device geolocation.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A deployer deploys Rule‑based activity recommender intended for AI that suggests simple outdoor activity ideas based on weather forecast and user preferences in the Leisure domain using User preferences, public weather forecast data.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A deployer deploys Forensic DNA reliability AI intended for AI system that evaluates the reliability of forensic DNA evidence in criminal cases in the Law enforcement domain using DNA sample metadata, laboratory quality metrics, chain‑of‑custody logs.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A provider deploys Legal fact‑extraction AI intended for AI system used by courts to automatically extract relevant facts from large document bundles and suggest legal arguments in the Administration of justice and democratic processes domain using Case files, statutes, prior judgments.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A provider deploys Template‑based itinerary generator intended for AI that creates basic travel itineraries for road trips based on user‑selected stops in the Travel domain using User stop list, public road map data.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A deployer deploys Event‑scale real‑time facial ID system intended for Deploy a stadium security AI that continuously matches faces of all attendees against a commercial database for marketing purposes in the Entertainment domain using Live video streams, facial feature vectors, ticket holder information, timestamps.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A provider deploys Reservation demand predictor intended for System that predicts short‑term demand for community‑center art‑exhibition spaces in the Culture domain using Public exhibition booking data, anonymised attendance records, local event calendars.", "gold": "LIMITED_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A deployer deploys Rule‑based flashcard creator intended for AI that suggests basic study flashcards for language vocabulary entered by the user in the Education domain using User word list, public bilingual dictionaries.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A deployer deploys Therapy emotion detection AI intended for Emotion recognition AI used in online therapy platforms to detect signs of anxiety and trigger therapist alerts in the Biometrics domain using Facial video, voice tone, interaction timing.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A deployer deploys Mental‑health emotion detection AI intended for Emotion recognition AI used in mental‑health apps to detect signs of depression and trigger alerts to clinicians in the Biometrics domain using Facial video, voice tone, interaction patterns.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A deployer deploys Rule‑based photography guide intended for AI that suggests basic photo composition tips based on user‑selected scene type in the Photography domain using User scene type, public composition rules.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A provider deploys Template‑based story generator intended for AI that generates short bedtime stories from user‑provided characters and setting in the Family entertainment domain using User characters and setting, public story templates.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A deployer deploys Admission decision AI intended for AI platform that determines admission to university programmes based on academic records and socio‑economic background in the Education and vocational training domain using Academic transcripts, standardized test scores, demographic data.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A provider deploys Usage prediction model intended for Tool that forecasts short‑term demand for public Wi‑Fi hotspots in city parks in the Public Services domain using Public hotspot usage statistics, anonymised connection logs, weather forecasts.", "gold": "LIMITED_RISK", "pred": "LIMITED_RISK", "ok": true }, { "item": "A provider deploys Phenology forecasting model intended for System that predicts optimal planting dates for hobby gardeners based on climate in the Home Gardening domain using Public climate data, horticultural calendars.", "gold": "LIMITED_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A provider deploys Civil‑service recruitment AI intended for AI platform that automates the selection of candidates for senior civil‑service positions, including predictive performance analytics in the Employment, workers’ management and access to self‑employment domain using Career histories, assessment scores, psychometric tests.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A deployer deploys Rule‑based reminder system intended for AI that offers basic car‑maintenance reminders based on vehicle model entered by the user in the Automotive (non‑critical) domain using User vehicle model, public maintenance schedule data.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A provider deploys Demand forecasting model intended for Tool that forecasts demand for seasonal agricultural equipment rentals in the Agriculture domain using Public agricultural activity statistics, weather forecasts, anonymised rental histories.", "gold": "LIMITED_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A deployer deploys CCTV facial data collection engine intended for Operate a retail analytics platform that builds a massive facial‑recognition database by extracting faces from CCTV feeds across shopping malls in the Retail domain using Video streams, timestamp data, location coordinates, shopper demographics.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A provider deploys Rule‑based conversational agent intended for AI-powered chatbot that answers general product FAQs on an e‑commerce site in the Customer service domain using User text queries, product catalogue, public FAQ documents.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A deployer deploys Rule‑based reminder system intended for AI that offers simple daily water‑intake reminders based on user‑selected weight and activity level in the Health & wellness domain using User weight and activity level, public hydration guidelines.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A provider deploys Real‑time library visitor profiling system intended for Offer a public‑library AI that scans visitors’ faces in real time to infer reading preferences and push targeted book recommendations in the Public services domain using Entrance camera video, facial embeddings, borrowing history, demographic data.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A deployer deploys Subliminal visual persuasion tool intended for Use a visual cue manipulation system that embeds subliminal images in online news feeds to sway political opinions in the Media domain using News article content, user click data, location data, political affiliation indicators.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A provider deploys Critical‑infrastructure grid management AI intended for AI system controlling the load balancing of a national electricity grid, acting as a safety component subject to the EU Electricity Directive conformity assessment in the Management and operation of critical infrastructure domain using Real‑time power demand, generation forecasts, sensor data from substations.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A provider deploys Acoustic prediction model intended for Tool that forecasts short‑term noise levels in city districts for residents in the Environment domain using Public traffic data, weather forecasts, anonymised noise sensor readings.", "gold": "LIMITED_RISK", "pred": "LIMITED_RISK", "ok": true }, { "item": "A deployer deploys Rule‑based scheduler intended for AI that offers simple indoor plant fertilisation schedule based on plant type entered by the user in the Home gardening domain using User plant type, public fertilisation guidelines.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A deployer deploys Civic engagement scoring tool intended for Run a transportation‑ticketing system that denies discounts to passengers with low civic engagement scores in the Transportation domain using Public transport usage, community event participation, social media activity, age.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A provider deploys Judicial decision‑support AI intended for AI assistant for judges that analyses case law and suggests legal reasoning, directly influencing judicial decisions in the Administration of justice and democratic processes domain using Legal texts, precedent databases, case facts.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A deployer deploys University dropout prediction AI intended for AI that predicts dropout risk for university students and automatically triggers remedial tutoring in the Education and vocational training domain using Academic records, attendance, socio‑economic data.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A provider deploys Template‑based tagline generator intended for AI that generates short product taglines from user‑provided product description in the Marketing domain using User product description, public marketing copy.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A provider deploys Grid load‑shedding AI intended for AI system that manages the load shedding of an electricity grid during peak demand, acting as a safety component in the Management and operation of critical infrastructure domain using Demand forecasts, generation capacity, real‑time consumption data.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A deployer deploys Rule‑based scheduler intended for AI that offers simple garden watering schedules based on plant type and climate zone in the Gardening domain using User plant type, public climate data.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A deployer deploys Exam stress detection AI intended for Emotion recognition AI used in online examinations to detect stress and possible cheating behaviour in the Biometrics domain using Webcam video, voice analysis, keystroke dynamics.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A deployer deploys Content generation model intended for AI‑enhanced personal language flashcard generator for vocabulary building in the Education domain using Public lexical databases, anonymised learner progress data.", "gold": "LIMITED_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A deployer deploys Recommendation engine intended for AI‑driven personal fitness coach that suggests workout plans based on user goals in the Health & Wellness domain using Anonymised activity logs, public exercise databases.", "gold": "LIMITED_RISK", "pred": "LIMITED_RISK", "ok": true }, { "item": "A deployer deploys Rule‑based advice system intended for AI that provides basic pet‑care tips based on animal type selected by the user in the Pet care domain using User animal type, public pet‑care guidelines.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A deployer deploys Asylum application analysis AI intended for AI system that assists officials in examining asylum applications by extracting relevant facts and scoring credibility in the Migration, asylum and border‑control management domain using Applicant statements, country‑of‑origin reports, supporting documents.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A provider deploys Template‑based subject line generator intended for AI that generates short email subject lines from user‑provided email content in the Productivity domain using User email body, public subject line corpus.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A deployer deploys Subliminal audio suggestion system intended for Deploy a smart‑home speaker that emits low‑volume cues to promote subscription to premium services in the Smart home domain using Voice command logs, household device usage, subscription status, household income.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A provider deploys Template‑based itinerary generator intended for AI that provides simple travel itinerary ideas based on destination and trip length in the Travel domain using User destination and duration, public travel guide data.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A deployer deploys Law‑enforcement profiling AI intended for AI system that profiles individuals for law‑enforcement investigations based on behavioural patterns in the Law enforcement domain using Social media activity, location history, prior convictions.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A provider deploys Social credit scoring AI intended for Provide a social‑scoring system that assigns credit scores based on citizens’ online behaviour and restricts access to public housing for low scores in the Public services domain using Social media activity, browsing history, geolocation data, public service usage.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A provider deploys Demand forecasting algorithm intended for System that predicts short‑term demand for municipal bike‑repair stations in the Urban Mobility domain using Public bike‑share usage data, anonymised repair logs, weather forecasts.", "gold": "LIMITED_RISK", "pred": "LIMITED_RISK", "ok": true }, { "item": "A provider deploys Digital forensic reliability AI intended for AI system that assists prosecutors in assessing the reliability of digital forensic evidence in the Law enforcement domain using File metadata, hash values, chain‑of‑custody documentation.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A provider deploys Anomaly detection on usage logs intended for Predictive maintenance scheduler for non‑critical office equipment (e.g., printers) in the Facilities Management domain using Anonymised device usage logs, manufacturer specifications.", "gold": "LIMITED_RISK", "pred": "LIMITED_RISK", "ok": true }, { "item": "A deployer deploys Rule‑based lighting advisor intended for AI that suggests basic indoor lighting colour temperature settings based on user‑selected mood in the Home improvement domain using User mood, public lighting colour guidelines.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A provider deploys Building fire‑suppression AI intended for AI safety component for automated fire‑suppression systems in high‑rise buildings, subject to EU construction product conformity assessment in the Product safety AI systems domain using Smoke detectors, temperature sensors, building layout data.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A provider deploys Rule‑based scheduler intended for System that predicts optimal cleaning schedules for office spaces based on occupancy patterns in the Facilities Management domain using Anonymised badge‑in data, public building usage statistics.", "gold": "LIMITED_RISK", "pred": "LIMITED_RISK", "ok": true }, { "item": "A deployer deploys Social security disability AI intended for AI that evaluates eligibility for social security disability benefits based on medical and employment data in the Access to and enjoyment of essential private services and essential public services and benefits domain using Medical reports, employment history, income statements.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A provider deploys Insurance pricing AI intended for AI that calculates life‑insurance premiums based on health data and lifestyle, affecting pricing in the Access to and enjoyment of essential private services and essential public services and benefits domain using Medical records, wearable device data, lifestyle questionnaires.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A provider deploys Real‑time shopper tracking platform intended for Offer a mall‑wide monitoring AI that uses remote biometric identification to track shoppers’ movements and sell heat‑map data to retailers in the Retail domain using CCTV video, facial embeddings, store layout maps, time data.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A provider deploys Time‑series demand predictor intended for Tool that forecasts demand for seasonal tourism services in coastal towns in the Tourism domain using Public tourism statistics, weather forecasts, event calendars.", "gold": "LIMITED_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A provider deploys Cross‑platform facial image harvesting system intended for Sell a social‑media monitoring tool that scrapes user‑uploaded photos to create a cross‑platform facial‑recognition repository for targeted advertising in the Advertising domain using User photos, hashtags, upload timestamps, device identifiers.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A deployer deploys Rule‑based flashcard creator intended for AI that offers basic language learning flashcards based on user‑chosen vocabulary in the Education domain using User vocabulary list, public bilingual word lists.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A provider deploys Social behaviour classification tool intended for Sell a recruitment AI that gives lower ranking to candidates with certain social media habits, affecting job offers in the Human resources domain using LinkedIn activity, personal posts, network size, location data.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A deployer deploys Video summarisation model intended for AI‑enhanced video summarisation for user‑generated content platforms in the Social Media domain using Publicly shared videos, anonymised metadata.", "gold": "LIMITED_RISK", "pred": "LIMITED_RISK", "ok": true }, { "item": "A deployer deploys Recommendation engine intended for AI‑based personal productivity assistant that suggests task prioritisation in the Consumer Technology domain using Anonymised task lists, public productivity research.", "gold": "LIMITED_RISK", "pred": "LIMITED_RISK", "ok": true }, { "item": "A provider deploys In‑store emotion detection system intended for Sell a retail‑store AI that analyzes shoppers’ micro‑expressions to gauge satisfaction and trigger upselling prompts in the Retail domain using CCTV facial video, purchase history, demographic data, time‑of‑day.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A deployer deploys Retrieval‑augmented generation model intended for AI‑enhanced chatbot for non‑critical IT help‑desk support in the IT Services domain using Public knowledge‑base articles, anonymised ticket logs.", "gold": "LIMITED_RISK", "pred": "LIMITED_RISK", "ok": true }, { "item": "A deployer deploys Workforce scheduling AI intended for AI system that assigns employees to shifts based on health data and performance metrics, affecting work conditions in the Employment, workers’ management and access to self‑employment domain using Health records, performance logs, availability calendars.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A provider deploys Template‑based joke generator intended for AI that generates short jokes from user‑provided topics in the Entertainment domain using User topics, public joke corpus.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A provider deploys Executive recruitment AI intended for AI system that automates the selection of candidates for senior management positions, including predictive performance modelling in the Employment, workers’ management and access to self‑employment domain using Leadership assessments, career history, psychometric data.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A provider deploys Rule‑based game idea generator intended for AI that suggests simple board game ideas for family gatherings in the Family entertainment domain using User age range, public game concept database.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A deployer deploys Real‑time biometric tracking engine intended for Run a city‑wide AI that uses real‑time remote biometric identification to monitor foot traffic and sell movement data to advertisers in the Smart city domain using Street‑level camera streams, facial embeddings, device IDs, time stamps.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A provider deploys Template‑based list generator intended for AI that creates simple travel packing lists from user‑selected destination and duration in the Travel domain using User destination and trip length, public packing guidelines.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A deployer deploys Demand forecasting model intended for AI system that predicts demand for non‑perishable goods in small retail stores in the Retail domain using Aggregated sales data, public holiday calendars, weather forecasts.", "gold": "LIMITED_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A provider deploys Subliminal visual cue engine intended for Provide a video streaming service that inserts barely perceptible visual prompts to increase subscription upgrades in the Entertainment domain using Video frames, user watch history, subscription status, demographic data.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A provider deploys Time‑series price predictor intended for System that predicts short‑term electricity price fluctuations for residential users in the Energy domain using Public market price data, weather forecasts, anonymised consumption patterns.", "gold": "LIMITED_RISK", "pred": "LIMITED_RISK", "ok": true }, { "item": "A provider deploys Sentiment aggregation platform intended for Tool that analyses public sentiment on product reviews to provide manufacturers with insights in the Manufacturing domain using Public product review texts, anonymised rating data.", "gold": "LIMITED_RISK", "pred": "LIMITED_RISK", "ok": true }, { "item": "A provider deploys Subconscious influence engine intended for Deploy a hidden‑audio influence platform that subtly nudges shoppers to purchase high‑margin products without their conscious awareness in the Retail domain using Audio streams, browsing history, purchase records, demographic profiles.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A provider deploys Economic vulnerability persuasion tool intended for Sell an AI‑driven charity donation app that manipulates low‑income families into over‑donating by highlighting emotional stories in the Non‑profit domain using Income data, donation history, demographic profiles, emotional content analysis.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A provider deploys Re‑offending risk AI intended for AI system used by police to assess the probability that a suspect will re‑offend, influencing parole decisions in the Law enforcement domain using Criminal history, psychological assessments, social environment data.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A deployer deploys Template‑based graphic generator intended for AI that generates decorative holiday greeting cards from user‑provided text in the Personalisation domain using User text, public holiday image assets.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A deployer deploys Rule‑based fitness planner intended for AI that suggests daily workout routines based on user‑selected fitness goals in the Health & fitness (non‑medical) domain using User‑selected goals, public exercise database.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A deployer deploys Scholarship allocation AI intended for AI that evaluates the suitability of candidates for scholarship programmes based on academic merit and socio‑economic background in the Education and vocational training domain using Grades, test scores, household income, demographic data.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A provider deploys Rule‑based optimizer intended for System that predicts optimal lighting schedules for office buildings to reduce energy use in the Facilities Management domain using Public daylight data, anonymised building occupancy patterns.", "gold": "LIMITED_RISK", "pred": "LIMITED_RISK", "ok": true }, { "item": "A deployer deploys Rule‑based strategy suggester intended for AI that suggests board game strategies for casual players in the Gaming domain using User game state description, public game rulebooks.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A deployer deploys Content generation model intended for AI‑based language learning flashcard generator for learners in the Education domain using Open linguistic corpora, anonymised learner progress data.", "gold": "LIMITED_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A provider deploys Rule‑based Q&A system intended for Chatbot that provides basic legal information on consumer rights (non‑advisory) in the Legal Services domain using Public consumer law texts, anonymised query logs.", "gold": "LIMITED_RISK", "pred": "LIMITED_RISK", "ok": true }, { "item": "A deployer deploys Nutrient estimation engine intended for AI‑enhanced personal recipe nutrition calculator for home cooks in the Health & Wellness domain using Public food composition tables, user‑submitted recipes.", "gold": "LIMITED_RISK", "pred": "LIMITED_RISK", "ok": true }, { "item": "A deployer deploys Speech‑to‑text model intended for AI‑enhanced video captioning tool for creators of non‑commercial content in the Content Creation domain using User‑uploaded video audio tracks, public language corpora.", "gold": "LIMITED_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A deployer deploys Recommendation engine intended for AI‑based personal habit‑formation coach that suggests incremental changes in the Consumer Technology domain using Anonymised habit logs, public behaviour change research.", "gold": "LIMITED_RISK", "pred": "LIMITED_RISK", "ok": true }, { "item": "A provider deploys Real‑time safety monitoring AI for medical robots intended for AI safety component for autonomous surgical robots that monitors instrument positioning and triggers emergency stop in the Product safety AI systems domain using Live video feed, force‑torque sensor data, patient vital signs.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A provider deploys Template‑based FAQ generator intended for AI that generates short FAQ entries from user‑provided product questions in the Customer service domain using User questions, public answer templates.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A provider deploys Educator emotion assessment system intended for Provide an AI that evaluates teachers’ emotional states during lectures to determine salary bonuses in the Education domain using Classroom video, audio tone, teacher ID, student feedback scores.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A provider deploys Drone traffic safety AI intended for AI safety component for autonomous drone traffic management, required to undergo conformity assessment under EU aviation regulations in the Product safety AI systems domain using Airspace data, drone telemetry, weather conditions.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A deployer deploys Urban facial data accumulation system intended for Deploy a smart‑city surveillance network that continuously captures and stores facial images from public streets to build a city‑wide facial‑recognition repository in the Smart city domain using Street‑level video feeds, timestamps, location coordinates, vehicle plate data.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A provider deploys Template‑based caption generator intended for AI that creates short social media captions from user‑provided image description in the Social media domain using User image description, public caption corpus.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A provider deploys Statistical analysis dashboard intended for Tool that analyses public transport punctuality trends to inform commuters in the Urban Mobility domain using Open transport schedule data, real‑time arrival feeds.", "gold": "LIMITED_RISK", "pred": "LIMITED_RISK", "ok": true }, { "item": "A provider deploys Statistical machine translation intended for AI that translates short phrases between common languages for casual conversation in the Communication domain using User‑provided text, public bilingual corpora.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A deployer deploys Exploitation‑aware loan persuasion system intended for Operate a loan‑approval chatbot that manipulates financially illiterate seniors into accepting predatory loan terms in the Finance domain using Credit history, age, financial literacy test results, income statements.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A provider deploys Template‑based signature generator intended for AI that generates short email signature blocks from user‑provided contact details in the Productivity domain using User contact details, public signature templates.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true } ], "meok_referee_off": [ { "item": "A provider deploys Behavioural loyalty scoring engine intended for Sell a consumer‑loyalty program that reduces reward tiers for users with undesirable online behaviour patterns in the Retail domain using Purchase history, online reviews, social media sentiment, demographic data.", "gold": "PROHIBITED", "pred": "HIGH_RISK", "ok": false }, { "item": "A provider deploys Predictive foot‑traffic model intended for Tool that forecasts foot‑traffic for retail malls to optimise shop‑front displays in the Retail domain using Public pedestrian count data, anonymised sales figures, weather forecasts.", "gold": "LIMITED_RISK", "pred": "LIMITED_RISK", "ok": true }, { "item": "A provider deploys Migration health‑risk assessment AI intended for AI that assesses health risks of migrants entering the EU, influencing border‑control decisions in the Migration, asylum and border‑control management domain using Medical records, travel history, epidemiological data.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A deployer deploys Real‑time driver biometric monitoring system intended for Deploy a city‑traffic AI that uses remote biometric identification to monitor drivers and sell driver‑profile data to insurance companies in the Transportation domain using Road‑side camera footage, facial images, vehicle registration data, timestamps.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A deployer deploys Student emotion analytics AI intended for Emotion recognition AI used in online learning platforms to gauge student engagement and adapt content delivery in the Biometrics domain using Webcam video, voice tone, interaction timestamps.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A provider deploys Collaborative filtering model intended for Recommendation system for online retailers that suggests complementary products without influencing purchase decisions in the E‑commerce domain using Aggregated, anonymised purchase histories, product metadata.", "gold": "LIMITED_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A provider deploys Template‑based tip generator intended for AI that creates short travel safety tips from user‑selected destination in the Travel domain using User destination, public travel safety advisories.", "gold": "MINIMAL_RISK", "pred": "LIMITED_RISK", "ok": false }, { "item": "A deployer deploys Image classification model intended for AI‑driven personal garden pest identifier that suggests non‑chemical remedies in the Home Gardening domain using Public pest image datasets, user‑uploaded garden photos.", "gold": "LIMITED_RISK", "pred": "LIMITED_RISK", "ok": true }, { "item": "A provider deploys Automated recruitment and candidate ranking AI intended for AI system used by a recruitment agency to filter and rank job applicants, including automated targeting of advertisements in the Employment, workers’ management and access to self‑employment domain using CVs, cover letters, online profiles, psychometric test results.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A deployer deploys Call‑centre emotion detection AI intended for Emotion recognition AI used in customer service call centres to detect frustration and trigger supervisor intervention in the Biometrics domain using Voice tone, speech content, interaction duration.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A provider deploys Emergency power‑grid management AI intended for AI system that manages the distribution of electricity during emergencies, automatically isolating faulty sections in the Management and operation of critical infrastructure domain using Grid sensor data, outage reports, load forecasts.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A deployer deploys Child benefit eligibility AI intended for AI that calculates eligibility for child benefit payments based on family income and composition in the Access to and enjoyment of essential private services and essential public services and benefits domain using Household income, number of dependents, tax records.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A provider deploys Theme‑park visitor tracking engine intended for Provide a real‑time facial recognition system for a theme park that tracks all guests to sell detailed movement patterns to third‑party marketers in the Entertainment domain using Live video streams, facial embeddings, ride usage logs, timestamps.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A deployer deploys Adaptive scheduling engine intended for AI‑driven personal study‑plan generator for students in the Education domain using Public curriculum outlines, anonymised student performance data.", "gold": "LIMITED_RISK", "pred": "LIMITED_RISK", "ok": true }, { "item": "A deployer deploys Statistical forecasting model intended for AI that provides weather forecasts for a specific city using public meteorological data in the Information services domain using Public weather APIs, location name.", "gold": "MINIMAL_RISK", "pred": "LIMITED_RISK", "ok": false }, { "item": "A provider deploys Template‑based summariser intended for AI that generates short product warranty summaries from user‑provided warranty text in the Consumer goods domain using User warranty text, public summarisation patterns.", "gold": "MINIMAL_RISK", "pred": "LIMITED_RISK", "ok": false }, { "item": "A provider deploys Chemical reactor emergency shutdown AI intended for AI safety component for automated emergency shutdown of industrial chemical reactors, subject to EU chemical safety conformity assessment in the Product safety AI systems domain using Temperature, pressure, gas concentration sensors.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A deployer deploys Rule‑based routine planner intended for AI that suggests basic daily stretch routines based on user‑selected flexibility level in the Wellbeing domain using User flexibility level, public stretch guide.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A provider deploys Partner violence risk assessment tool intended for Offer a dating‑app AI that blocks users predicted to become violent partners based on personality traits inferred from profile text in the Online dating domain using Profile descriptions, message content, psycholinguistic analysis, age.", "gold": "PROHIBITED", "pred": "HIGH_RISK", "ok": false }, { "item": "A deployer deploys Pattern‑recognition model intended for AI‑based personal sleep‑quality coach that offers non‑medical advice in the Health & Wellness domain using Anonymised sleep sensor data, public sleep research findings.", "gold": "LIMITED_RISK", "pred": "LIMITED_RISK", "ok": true }, { "item": "A deployer deploys Crowd density biometric monitoring system intended for Deploy a public‑square AI that uses remote biometric identification to monitor crowd density and sell real‑time occupancy data to event organizers in the Public spaces domain using CCTV video, facial detection data, location coordinates, time stamps.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A provider deploys Demand forecasting algorithm intended for System that predicts short‑term demand for public bike‑sharing stations in the Urban Mobility domain using Public bike‑share usage data, weather forecasts, event calendars.", "gold": "LIMITED_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A deployer deploys Rule‑based advice system intended for AI that offers simple DIY home repair tips based on user‑selected problem description in the Home improvement domain using User problem description, public repair guides.", "gold": "MINIMAL_RISK", "pred": "LIMITED_RISK", "ok": false }, { "item": "A provider deploys Time‑series forecasting model intended for Tool that forecasts agricultural yield for small farms to aid planning, not used for insurance underwriting in the Agriculture domain using Open satellite imagery, weather forecasts, anonymised farm production records.", "gold": "LIMITED_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A deployer deploys Retail mood‑adaptive AI intended for Emotion recognition AI used in retail environments to adapt lighting and music based on shopper mood in the Biometrics domain using Facial video, voice tone, movement patterns.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A deployer deploys Rule‑based routing assistant intended for AI that suggests simple bike route alternatives based on user‑selected start and end points in the Transportation (non‑critical) domain using User locations, public map data.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A deployer deploys Generative language model with limited scope intended for AI assistant that drafts informal email replies for personal use in the Personal productivity domain using User‑provided email text, public language corpora.", "gold": "MINIMAL_RISK", "pred": "LIMITED_RISK", "ok": false }, { "item": "A provider deploys Gas pipeline safety AI intended for AI safety component for autonomous gas pipeline monitoring that detects leaks and initiates shutdown in the Product safety AI systems domain using Pressure sensors, acoustic emission data, temperature readings.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A provider deploys Occupancy prediction model intended for Tool that forecasts short‑term demand for public parking spaces in city centres in the Urban Mobility domain using Public parking sensor data, anonymised historical occupancy, event calendars.", "gold": "LIMITED_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A deployer deploys Pattern‑recognition dashboard intended for AI‑enhanced personal habit‑tracking app that visualises weekly patterns in the Consumer Technology domain using Anonymised habit logs, public habit‑formation studies.", "gold": "LIMITED_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A deployer deploys Rule‑based lighting advisor intended for AI that suggests simple indoor lighting setups based on room size entered by the user in the Home improvement domain using User room dimensions, public lighting guidelines.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A provider deploys Vulnerability targeting advertising platform intended for Offer a targeted advertising AI that exploits the cognitive vulnerabilities of users with dementia to sell unnecessary health supplements in the Healthcare domain using Medical records, cognitive assessment scores, browsing history, purchase logs.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A provider deploys Demand forecasting algorithm intended for System that predicts short‑term demand for municipal recycling collection services in the Environment domain using Public waste generation statistics, anonymised household recycling data, weather forecasts.", "gold": "LIMITED_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A deployer deploys Referendum influence AI intended for AI that manipulates online news feeds to sway public opinion ahead of a referendum in the Administration of justice and democratic processes domain using User browsing history, sentiment analysis, content recommendation algorithms.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A provider deploys Government recruitment AI intended for AI platform that automates the screening of job applications for a government agency, including bias‑checking in the Employment, workers’ management and access to self‑employment domain using Applicant CVs, questionnaire responses, background checks.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A provider deploys Mass surveillance facial identification platform intended for Offer a real‑time crowd monitoring system that continuously scans public squares with facial recognition to identify and track all individuals for commercial profiling in the Public spaces domain using Live video feeds, facial image database, timestamp, location coordinates.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A deployer deploys Unemployment benefit eligibility AI intended for AI that determines eligibility for unemployment benefits based on income and work history, with automatic benefit allocation in the Access to and enjoyment of essential private services and essential public services and benefits domain using Tax records, employment contracts, bank statements.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A provider deploys Template‑based table generator intended for AI that generates short product comparison tables from user‑provided feature lists in the E‑commerce domain using User feature lists, public comparison formats.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A provider deploys Credit scoring AI intended for AI that calculates credit scores for loan applicants, influencing creditworthiness decisions in the Access to and enjoyment of essential private services and essential public services and benefits domain using Financial history, transaction data, employment records.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A provider deploys Template‑based infographic generator intended for AI that creates basic infographic layouts from user‑provided statistics in the Data visualisation domain using User statistics, public icon sets.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A provider deploys Residential heating safety AI intended for AI safety component for automated gas valve control in residential heating systems, subject to EU gas appliance conformity assessment in the Product safety AI systems domain using Temperature sensors, pressure sensors, user settings.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A deployer deploys Goal‑recommendation engine intended for AI‑based personal finance goal tracker that suggests saving milestones in the FinTech domain using Anonymised financial transaction categories, public cost‑of‑living data.", "gold": "LIMITED_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A deployer deploys Speech analysis model intended for AI‑driven personal language pronunciation trainer for learners in the Education domain using Public phonetic corpora, anonymised user recordings.", "gold": "LIMITED_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A deployer deploys Rule‑based financial advisor intended for AI that offers basic budgeting tips based on user‑entered expense categories in the Personal finance domain using User expense categories, public budgeting guidelines.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A deployer deploys Rule‑based activity suggester intended for AI that suggests basic study break activities based on user‑selected subject and time constraints in the Education domain using User subject and time, public break activity ideas.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A deployer deploys Social interaction based credit assessment system intended for Deploy a banking AI that reduces loan limits for users with low social engagement scores in the Banking domain using Social media engagement, transaction history, credit score, age.", "gold": "PROHIBITED", "pred": "HIGH_RISK", "ok": false }, { "item": "A deployer deploys Neural machine translation model intended for AI‑based language translation service for travel documents in the Travel domain using Public multilingual corpora, anonymised user inputs.", "gold": "LIMITED_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A deployer deploys Rule‑based recommendation engine intended for AI‑driven personal finance budgeting assistant that suggests savings plans in the FinTech domain using Anonymised transaction categories, public cost‑of‑living indices.", "gold": "LIMITED_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A provider deploys Template‑based sticker generator intended for AI that creates simple animated stickers from user‑provided text and emojis in the Social media domain using User text and emojis, public sticker templates.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A deployer deploys Conversational AI with predefined answer templates intended for Chatbot that assists citizens with non‑urgent municipal service queries (e.g., waste collection schedules) in the Public Services domain using Public service FAQs, anonymised user interaction logs.", "gold": "LIMITED_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A deployer deploys Generative layout model intended for AI‑enhanced personal garden layout planner that suggests plant placement in the Home Gardening domain using Public plant hardiness data, user‑provided garden dimensions.", "gold": "LIMITED_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A deployer deploys Recommendation engine intended for AI‑driven virtual career mentor that suggests skill‑development pathways in the Career Services domain using Public occupational standards, anonymised user skill profiles.", "gold": "LIMITED_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A deployer deploys Workout video facial data collector intended for Run a fitness‑app that collects users’ facial images during workout recordings to enlarge a facial‑recognition dataset for health‑related advertising in the Health & fitness domain using Video recordings, heart‑rate data, user demographics, location data.", "gold": "PROHIBITED", "pred": "HIGH_RISK", "ok": false }, { "item": "A deployer deploys Evidence reliability AI intended for AI system that evaluates the reliability of digital evidence in criminal investigations in the Law enforcement domain using Metadata of digital files, chain‑of‑custody logs, forensic analysis results.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A provider deploys Covert conversational influence AI intended for Sell a chatbot that uses covert language patterns to steer users toward purchasing insurance policies in the Insurance domain using Chat logs, user age, income level, previous insurance claims.", "gold": "PROHIBITED", "pred": "HIGH_RISK", "ok": false }, { "item": "A deployer deploys Behavioural evaluation and scoring system intended for Implement a city‑wide citizen rating platform that limits eligibility for municipal permits based on a behavioural score in the Local government domain using Public transport usage, waste disposal records, online interactions, demographic data.", "gold": "PROHIBITED", "pred": "HIGH_RISK", "ok": false }, { "item": "A deployer deploys Rule‑based pattern recommender intended for AI that suggests simple knitting patterns based on yarn weight selected by the user in the Crafts domain using User yarn weight, public knitting pattern database.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A deployer deploys Doctoral scholarship allocation AI intended for AI that evaluates the suitability of candidates for doctoral scholarships based on research output and academic record in the Education and vocational training domain using Publications, citation counts, grades, demographic data.", "gold": "HIGH_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A provider deploys Template‑based caption generator intended for AI that creates short social media story captions from user‑provided event description in the Social media domain using User event description, public story caption corpus.", "gold": "MINIMAL_RISK", "pred": "HIGH_RISK", "ok": false }, { "item": "A provider deploys Transit demand predictor intended for System that predicts optimal bus frequencies for non‑peak hours to reduce emissions in the Urban Mobility domain using Public ridership data, anonymised ticket sales, weather forecasts.", "gold": "LIMITED_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A provider deploys Content‑based recommendation system intended for AI that recommends books based on genre preferences entered by the user in the Literature domain using User‑selected genres, public book metadata.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A deployer deploys Public assistance eligibility AI intended for AI tool that evaluates eligibility for social housing benefits, calculating entitlement and potential revocation in the Access to and enjoyment of essential private services and essential public services and benefits domain using Income statements, household composition, rental market data.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A deployer deploys Academic vulnerability persuasion engine intended for Run an AI‑powered education portal that manipulates low‑performing students into purchasing expensive tutoring packages in the Education domain using Test scores, socioeconomic background, device usage, learning style data.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A provider deploys Trade‑union membership inference engine intended for Provide a consumer‑insights platform that groups shoppers by inferred trade‑union membership from facial cues for price discrimination in the Retail domain using In‑store camera images, purchase logs, demographic data, loyalty card ID.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A deployer deploys Rule‑based placement advisor intended for AI that suggests simple indoor plant placement based on light level entered by the user in the Home gardening domain using User light level, public plant light requirements.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A provider deploys Generative text model with limited scope intended for AI that creates simple meme captions from user‑provided images in the Social media domain using User images, public meme text corpus.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A deployer deploys Religious belief biometric classifier intended for Run a recruitment AI that filters candidates based on inferred religious beliefs derived from facial features in the Human resources domain using Video interview recordings, facial landmark data, CV data, location.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A provider deploys Recidivism risk assessment AI intended for AI tool that predicts recidivism risk for parole decisions, influencing release in the Law enforcement domain using Criminal record, behavioural reports, socio‑economic data.", "gold": "HIGH_RISK", "pred": "PROHIBITED", "ok": false }, { "item": "A provider deploys Template‑based mood‑board generator intended for AI that creates simple visual mood boards from user‑provided colour palettes in the Design domain using User colour palette, public image assets.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A provider deploys Template‑based quote generator intended for AI that generates short motivational quotes from user‑provided keywords in the Personal development domain using User keywords, public quote database.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A deployer deploys Generative design model intended for AI‑based virtual interior designer that suggests furniture layouts in the Home Improvement domain using Public furniture catalogs, anonymised room dimensions.", "gold": "LIMITED_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A provider deploys Witness credibility AI intended for AI system that assists police in assessing the credibility of witness statements using linguistic analysis in the Law enforcement domain using Transcripts, voice recordings, prior statements.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A deployer deploys Educational emotion inference engine intended for Deploy an AI that monitors students’ facial expressions during online classes to infer engagement levels and assign grades in the Education domain using Webcam video, audio tone, interaction timestamps, student identifiers.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A deployer deploys Rule‑based workout planner intended for AI that suggests simple indoor workout routines based on user‑selected equipment availability in the Fitness domain using User equipment list, public exercise database.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A deployer deploys Racial segmentation profiling system intended for Deploy an AI that analyses customers’ facial characteristics to assign them to racial market segments for product pricing in the Retail domain using Store camera footage, purchase history, demographic data, location.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A provider deploys Attendance predictor intended for Tool that forecasts short‑term demand for community‑center workshops in the Community Services domain using Public event calendars, anonymised past attendance records, weather forecasts.", "gold": "LIMITED_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A provider deploys Student criminal risk prediction engine intended for Provide a school‑discipline AI that predicts which students are likely to commit violent acts using personality profiling in the Education domain using Student essays, social media activity, behavioural logs, psychometric test results.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A deployer deploys Subliminal emoji insertion tool intended for Run a social media filter that embeds subliminal emojis to increase engagement with sponsored posts in the Social media domain using Post content, user interaction data, follower demographics, ad targeting parameters.", "gold": "PROHIBITED", "pred": "HIGH_RISK", "ok": false }, { "item": "A provider deploys Gas network pressure control AI intended for AI system that controls the pressure valves of a natural‑gas distribution network to prevent over‑pressurisation in the Management and operation of critical infrastructure domain using Pressure sensor data, flow rates, demand forecasts.", "gold": "HIGH_RISK", "pred": "LIMITED_RISK", "ok": false }, { "item": "A provider deploys Puzzle‑generation algorithm intended for AI that creates simple crossword puzzles from user‑provided word lists in the Education (recreational) domain using User word list, public crossword patterns.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A provider deploys Instantaneous facial ID advertising engine intended for Provide a real‑time facial recognition system for a shopping centre that flags all customers for personalized ad delivery without consent in the Retail domain using Live video streams, facial templates, purchase history, demographic profiles.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A deployer deploys Rule‑based audio guide generator intended for AI that offers basic meditation guidance based on user‑selected duration in the Wellbeing domain using User duration choice, public meditation scripts.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A provider deploys Usage forecasting model intended for System that predicts short‑term demand for community‑center sports facilities in the Community Services domain using Public facility booking data, anonymised attendance records, weather forecasts.", "gold": "LIMITED_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A deployer deploys Event‑based facial data collector intended for Run an event‑management AI that collects attendee facial images from ticket scans to expand a commercial facial‑recognition dataset in the Event services domain using Ticket scan images, attendee names, event location, timestamps.", "gold": "PROHIBITED", "pred": "HIGH_RISK", "ok": false }, { "item": "A provider deploys Transit emotion detection platform intended for Offer a public‑transport monitoring AI that infers commuter stress levels to adjust ticket pricing dynamically in the Transportation domain using Onboard camera footage, ambient noise, passenger IDs, ticket data.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A deployer deploys Nutritional analysis engine intended for AI‑based recipe nutrition analyser that provides macro‑nutrient breakdowns in the Health & Wellness domain using Public food composition databases, user‑submitted recipes.", "gold": "LIMITED_RISK", "pred": "HIGH_RISK", "ok": false }, { "item": "A provider deploys Route optimisation algorithm intended for System that predicts waste‑collection truck routes to improve efficiency without affecting emergency services in the Waste Management domain using Public road network data, anonymised waste‑generation statistics, traffic forecasts.", "gold": "LIMITED_RISK", "pred": "LIMITED_RISK", "ok": true }, { "item": "A provider deploys Civic participation scoring system intended for Provide a public‑service access system that blocks low‑scoring citizens from voting assistance tools in the Public administration domain using Voting history, online political activity, demographic data, location.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A provider deploys Personality‑based fraud risk predictor intended for Sell a credit‑card fraud detection system that denies credit to users predicted to commit fraud based on personality analysis in the Finance domain using Transaction history, social media sentiment, psychometric scores, location data.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A deployer deploys Adaptive learning outcome AI intended for AI that evaluates student learning outcomes and automatically adjusts curriculum pathways in the Education and vocational training domain using Assessment scores, interaction logs, learning analytics.", "gold": "HIGH_RISK", "pred": "LIMITED_RISK", "ok": false }, { "item": "A deployer deploys Template recommendation engine intended for AI‑enhanced resume formatting assistant that suggests layout improvements in the Career Services domain using Public resume templates, anonymised user drafts.", "gold": "LIMITED_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A deployer deploys Emergency medical assistance eligibility AI intended for AI that determines eligibility for emergency medical assistance subsidies based on income and health status in the Access to and enjoyment of essential private services and essential public services and benefits domain using Medical records, income statements, household composition.", "gold": "HIGH_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A deployer deploys Rule‑based meal planner intended for AI that suggests basic meal planning ideas based on user‑selected dietary preferences in the Food & nutrition domain using User dietary preferences, public recipe database.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A deployer deploys Facial micro‑expression crime risk assessor intended for Run an employment screening tool that estimates future criminal behaviour of applicants based on facial micro‑expressions in the Human resources domain using Interview video recordings, facial landmark data, demographic information, prior convictions.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A deployer deploys Content‑based recommendation engine intended for AI‑based itinerary planner for tourists that suggests attractions based on preferences in the Tourism domain using Public points‑of‑interest data, anonymised user preference surveys.", "gold": "LIMITED_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A deployer deploys Pattern‑recognition dashboard intended for AI‑driven personal habit‑tracker that visualises daily routines in the Consumer Technology domain using Anonymised user activity logs, public habit‑formation research.", "gold": "LIMITED_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A deployer deploys Employee performance monitoring AI intended for AI system that monitors employee productivity through webcam and keystroke analysis, influencing performance evaluations in the Employment, workers’ management and access to self‑employment domain using Video feeds, keyboard/mouse activity, task completion timestamps.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A provider deploys Exam monitoring AI intended for AI platform that evaluates student exam submissions for plagiarism and cheating in real time in the Education and vocational training domain using Student code submissions, text similarity databases, webcam feeds.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A deployer deploys Rule‑based decor advisor intended for AI that offers basic home‑decor suggestions using user‑selected colour palettes in the Home improvement domain using User colour choices, public interior design image set.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A provider deploys Real‑time passenger identification system intended for Provide a transport hub AI that identifies every traveler in real time to build behavioural profiles for targeted ticket offers in the Transportation domain using CCTV facial video, ticket scan data, travel itinerary, demographic data.", "gold": "PROHIBITED", "pred": "HIGH_RISK", "ok": false }, { "item": "A deployer deploys Content recommendation engine intended for AI‑based virtual book club facilitator that suggests discussion topics in the Publishing domain using Public book summaries, anonymised user reading preferences.", "gold": "LIMITED_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A provider deploys Template‑based video generator intended for AI that creates simple birthday greeting videos from user‑provided photos and text in the Personalisation domain using User photos and text, public video templates.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A provider deploys Image style‑transfer model intended for AI that creates simple cartoon avatars from user‑drawn sketches in the Creative tools domain using User sketch images, publicly available cartoon style datasets.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A deployer deploys Extractive summarisation model intended for AI‑enhanced document summarisation tool for corporate reports in the Business Services domain using Corporate documents (non‑personal), publicly available reports.", "gold": "LIMITED_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A deployer deploys Rule‑based routine planner intended for AI that suggests simple bedtime routines based on user‑selected sleep goals in the Health & wellness domain using User sleep goal, public sleep hygiene tips.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A provider deploys Public‑sector recruitment AI intended for AI system that automates the selection of candidates for public sector jobs, including ranking and shortlisting in the Employment, workers’ management and access to self‑employment domain using CVs, application forms, competency test results.", "gold": "HIGH_RISK", "pred": "LIMITED_RISK", "ok": false }, { "item": "A provider deploys Template‑based recipe generator intended for AI that generates cooking recipes from a list of ingredients supplied by the user in the Lifestyle domain using User‑provided ingredient list, public recipe database.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A provider deploys Template‑based title generator intended for AI that generates short podcast episode titles from user‑provided topic description in the Media domain using User topic description, public podcast title corpus.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A provider deploys Statistical reporting dashboard intended for Tool that analyses public transport punctuality to generate performance reports for city planners in the Urban Mobility domain using Open transport schedule data, real‑time arrival feeds.", "gold": "LIMITED_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A provider deploys Biometric categorisation AI intended for AI system that categorises individuals by ethnicity based on facial features for law‑enforcement profiling in the Biometrics domain using High‑resolution facial images, demographic databases.", "gold": "HIGH_RISK", "pred": "PROHIBITED", "ok": false }, { "item": "A deployer deploys Multi‑objective optimization engine intended for AI‑driven personal travel itinerary optimizer that balances cost and time in the Travel domain using Public transport timetables, flight price APIs, anonymised user preferences.", "gold": "LIMITED_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A deployer deploys Aggression inference risk scoring system intended for Deploy a neighborhood safety platform that flags residents as high‑risk based on inferred aggression levels from online comments in the Public safety domain using Forum comments, video uploads, facial expression analysis, demographic data.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A deployer deploys Rule‑based scheduler intended for AI that creates basic study schedules based on user‑selected subjects and available hours in the Education domain using User subjects and time availability, public study‑tip resources.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A deployer deploys Migration security profiling AI intended for AI that profiles migrants based on travel routes and biometric data to assess security risk in the Migration, asylum and border‑control management domain using Passport scans, travel itineraries, facial images.", "gold": "HIGH_RISK", "pred": "PROHIBITED", "ok": false }, { "item": "A deployer deploys Subliminal visual influence system intended for Operate an AI‑driven billboard network that flashes ultra‑fast images to influence driver route choices in the Transportation domain using Traffic camera feeds, driver location data, advertisement inventory, vehicle type.", "gold": "PROHIBITED", "pred": "HIGH_RISK", "ok": false }, { "item": "A provider deploys Subconscious audio persuasion module intended for Launch a gaming platform that uses hidden sound frequencies to encourage in‑game purchases in the Gaming domain using Game audio streams, player purchase history, session duration, age verification data.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A deployer deploys Online behaviour crime risk evaluator intended for Implement a public‑housing allocation AI that excludes applicants deemed likely to engage in illegal activities based on online behaviour in the Housing domain using Search queries, forum participation, social media posts, income data.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A provider deploys Rule‑based music selector intended for AI that generates basic workout playlists matching user‑selected tempo in the Fitness domain using User tempo preference, public music metadata.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A provider deploys Rule‑based project recommender intended for AI that suggests simple DIY craft projects based on materials the user has at home in the Hobbies domain using User‑listed materials, public craft instructions.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A provider deploys Template‑based card generator intended for AI that creates basic greeting card designs from user‑provided occasion and message in the Personalisation domain using User occasion and message, public design assets.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A deployer deploys Asylum‑seeker prioritisation AI intended for AI that profiles asylum seekers based on travel routes, health status and demographic data to prioritize processing in the Migration, asylum and border‑control management domain using Travel documents, medical records, demographic attributes.", "gold": "HIGH_RISK", "pred": "PROHIBITED", "ok": false }, { "item": "A provider deploys User‑video facial feature extractor intended for Provide an AI that extracts facial features from user‑generated video content to create a commercial biometric database for product personalization in the E‑commerce domain using Uploaded videos, frame timestamps, user profile data, device metadata.", "gold": "PROHIBITED", "pred": "HIGH_RISK", "ok": false }, { "item": "A deployer deploys Rule‑based horticulture advisor intended for AI that suggests garden plant arrangements based on sunlight exposure entered by the user in the Gardening domain using User‑provided garden parameters, public plant database.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A deployer deploys Political persuasion AI intended for AI system that predicts the likelihood of a citizen voting for a particular party and tailors political advertising accordingly in the Administration of justice and democratic processes domain using Voting history, social media activity, demographic data.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A provider deploys Demand forecasting algorithm intended for Tool that forecasts demand for public library books to aid acquisition planning in the Public Services domain using Public library circulation data, anonymised patron demographics, event calendars.", "gold": "LIMITED_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A provider deploys Template‑based outline generator intended for AI that creates basic slide deck outlines from user‑provided presentation topic in the Productivity domain using User topic, public slide structure templates.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A deployer deploys Computer‑vision classifier intended for AI‑assisted image tagging for personal photo libraries in the Consumer Technology domain using User‑uploaded images (non‑personal identifiers), public image datasets.", "gold": "LIMITED_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A deployer deploys Asylum claim outcome prediction AI intended for AI system that profiles asylum seekers to predict likelihood of successful claim, influencing decision making in the Migration, asylum and border‑control management domain using Personal narratives, country‑of‑origin risk data, health records.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A provider deploys Industrial fire‑detection AI intended for AI safety component for automated fire‑detection and sprinkler activation in industrial facilities, subject to EU machinery conformity assessment in the Product safety AI systems domain using Smoke detectors, temperature sensors, airflow data.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A provider deploys Automotive emergency braking AI intended for AI safety component for autonomous vehicle braking systems, required to undergo EU type‑approval conformity assessment in the Product safety AI systems domain using Lidar, radar, camera feeds, vehicle speed.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A provider deploys Emergency call classification AI intended for AI system that assists courts in classifying emergency calls and prioritising dispatch of first responders in the Access to and enjoyment of essential private services and essential public services and benefits domain using Call audio, caller location, incident description.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A provider deploys Predictive routing engine using anonymised traffic and weather data intended for AI tool that suggests optimal delivery routes for small‑business couriers, avoiding emergency or critical‑infrastructure corridors in the Logistics domain using Public traffic flow data, open‑source weather forecasts, anonymised GPS traces.", "gold": "LIMITED_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A deployer deploys Rule‑based optimizer using consumption patterns intended for AI‑driven energy‑consumption advisory for residential users, offering tips to reduce bills in the Energy domain using Anonymised household electricity usage statistics, public weather data.", "gold": "LIMITED_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A provider deploys Template‑based GIF generator intended for AI that creates simple animated GIFs from user‑provided short video clips in the Social media domain using User video clip, public animation presets.", "gold": "MINIMAL_RISK", "pred": "HIGH_RISK", "ok": false }, { "item": "A provider deploys Adaptive learning engine intended for Virtual tutor that offers personalized math practice problems for secondary students in the Education domain using Open educational resources, anonymised student performance data.", "gold": "LIMITED_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A provider deploys Victim risk assessment AI intended for AI system used by police to assess the risk of a person becoming a victim of violent crime, influencing resource allocation in the Law enforcement domain using Crime statistics, personal history, location data, social network analysis.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A provider deploys Education placement AI intended for AI that determines the appropriate level of education for students based on prior achievement and aptitude tests in the Education and vocational training domain using Test scores, learning analytics, socio‑demographic data.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A provider deploys Template‑based note generator intended for AI that generates short thank‑you note drafts from user‑provided occasion details in the Personal communication domain using User occasion details, public thank‑you note corpus.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A deployer deploys Rule‑based habit recommender intended for AI that offers simple daily habit suggestions based on user‑selected improvement area in the Personal development domain using User improvement area, public habit‑building tips.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A provider deploys Template‑based logo generator intended for AI that generates simple logo concepts from user‑provided brand keywords in the Design domain using User brand keywords, public icon library.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A provider deploys Behavioral audio influence platform intended for Offer a background music AI that embeds low‑frequency tones to increase impulse buying in supermarkets in the Supermarket domain using In‑store audio feeds, sales data, shopper movement patterns, age data.", "gold": "PROHIBITED", "pred": "HIGH_RISK", "ok": false }, { "item": "A provider deploys Template‑based description generator intended for AI that generates short product description snippets from user‑provided feature list in the E‑commerce domain using User feature list, public marketing copy.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A provider deploys Template‑based summariser intended for AI that generates short product review summaries from user‑provided review texts in the E‑commerce domain using User review texts, public summarisation patterns.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A deployer deploys Social influence evaluation platform intended for Implement a university admissions AI that lowers acceptance chances for applicants with low social media influence scores in the Education domain using Follower counts, post engagement, extracurricular activity logs, academic records.", "gold": "PROHIBITED", "pred": "HIGH_RISK", "ok": false }, { "item": "A deployer deploys Rule‑based story suggester intended for AI that suggests basic bedtime story ideas based on child’s age entered by the user in the Family entertainment domain using User child age, public story templates.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A deployer deploys Social‑media profiling AI intended for AI that profiles individuals for law‑enforcement investigations based on social media activity and location data in the Law enforcement domain using Posts, likes, check‑ins, device geolocation.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A deployer deploys Rule‑based activity recommender intended for AI that suggests simple outdoor activity ideas based on weather forecast and user preferences in the Leisure domain using User preferences, public weather forecast data.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A deployer deploys Forensic DNA reliability AI intended for AI system that evaluates the reliability of forensic DNA evidence in criminal cases in the Law enforcement domain using DNA sample metadata, laboratory quality metrics, chain‑of‑custody logs.", "gold": "HIGH_RISK", "pred": "LIMITED_RISK", "ok": false }, { "item": "A provider deploys Legal fact‑extraction AI intended for AI system used by courts to automatically extract relevant facts from large document bundles and suggest legal arguments in the Administration of justice and democratic processes domain using Case files, statutes, prior judgments.", "gold": "HIGH_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A provider deploys Template‑based itinerary generator intended for AI that creates basic travel itineraries for road trips based on user‑selected stops in the Travel domain using User stop list, public road map data.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A deployer deploys Event‑scale real‑time facial ID system intended for Deploy a stadium security AI that continuously matches faces of all attendees against a commercial database for marketing purposes in the Entertainment domain using Live video streams, facial feature vectors, ticket holder information, timestamps.", "gold": "PROHIBITED", "pred": "HIGH_RISK", "ok": false }, { "item": "A provider deploys Reservation demand predictor intended for System that predicts short‑term demand for community‑center art‑exhibition spaces in the Culture domain using Public exhibition booking data, anonymised attendance records, local event calendars.", "gold": "LIMITED_RISK", "pred": "LIMITED_RISK", "ok": true }, { "item": "A deployer deploys Rule‑based flashcard creator intended for AI that suggests basic study flashcards for language vocabulary entered by the user in the Education domain using User word list, public bilingual dictionaries.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A deployer deploys Therapy emotion detection AI intended for Emotion recognition AI used in online therapy platforms to detect signs of anxiety and trigger therapist alerts in the Biometrics domain using Facial video, voice tone, interaction timing.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A deployer deploys Mental‑health emotion detection AI intended for Emotion recognition AI used in mental‑health apps to detect signs of depression and trigger alerts to clinicians in the Biometrics domain using Facial video, voice tone, interaction patterns.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A deployer deploys Rule‑based photography guide intended for AI that suggests basic photo composition tips based on user‑selected scene type in the Photography domain using User scene type, public composition rules.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A provider deploys Template‑based story generator intended for AI that generates short bedtime stories from user‑provided characters and setting in the Family entertainment domain using User characters and setting, public story templates.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A deployer deploys Admission decision AI intended for AI platform that determines admission to university programmes based on academic records and socio‑economic background in the Education and vocational training domain using Academic transcripts, standardized test scores, demographic data.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A provider deploys Usage prediction model intended for Tool that forecasts short‑term demand for public Wi‑Fi hotspots in city parks in the Public Services domain using Public hotspot usage statistics, anonymised connection logs, weather forecasts.", "gold": "LIMITED_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A provider deploys Phenology forecasting model intended for System that predicts optimal planting dates for hobby gardeners based on climate in the Home Gardening domain using Public climate data, horticultural calendars.", "gold": "LIMITED_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A provider deploys Civil‑service recruitment AI intended for AI platform that automates the selection of candidates for senior civil‑service positions, including predictive performance analytics in the Employment, workers’ management and access to self‑employment domain using Career histories, assessment scores, psychometric tests.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A deployer deploys Rule‑based reminder system intended for AI that offers basic car‑maintenance reminders based on vehicle model entered by the user in the Automotive (non‑critical) domain using User vehicle model, public maintenance schedule data.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A provider deploys Demand forecasting model intended for Tool that forecasts demand for seasonal agricultural equipment rentals in the Agriculture domain using Public agricultural activity statistics, weather forecasts, anonymised rental histories.", "gold": "LIMITED_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A deployer deploys CCTV facial data collection engine intended for Operate a retail analytics platform that builds a massive facial‑recognition database by extracting faces from CCTV feeds across shopping malls in the Retail domain using Video streams, timestamp data, location coordinates, shopper demographics.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A provider deploys Rule‑based conversational agent intended for AI-powered chatbot that answers general product FAQs on an e‑commerce site in the Customer service domain using User text queries, product catalogue, public FAQ documents.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A deployer deploys Rule‑based reminder system intended for AI that offers simple daily water‑intake reminders based on user‑selected weight and activity level in the Health & wellness domain using User weight and activity level, public hydration guidelines.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A provider deploys Real‑time library visitor profiling system intended for Offer a public‑library AI that scans visitors’ faces in real time to infer reading preferences and push targeted book recommendations in the Public services domain using Entrance camera video, facial embeddings, borrowing history, demographic data.", "gold": "PROHIBITED", "pred": "HIGH_RISK", "ok": false }, { "item": "A deployer deploys Subliminal visual persuasion tool intended for Use a visual cue manipulation system that embeds subliminal images in online news feeds to sway political opinions in the Media domain using News article content, user click data, location data, political affiliation indicators.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A provider deploys Critical‑infrastructure grid management AI intended for AI system controlling the load balancing of a national electricity grid, acting as a safety component subject to the EU Electricity Directive conformity assessment in the Management and operation of critical infrastructure domain using Real‑time power demand, generation forecasts, sensor data from substations.", "gold": "HIGH_RISK", "pred": "LIMITED_RISK", "ok": false }, { "item": "A provider deploys Acoustic prediction model intended for Tool that forecasts short‑term noise levels in city districts for residents in the Environment domain using Public traffic data, weather forecasts, anonymised noise sensor readings.", "gold": "LIMITED_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A deployer deploys Rule‑based scheduler intended for AI that offers simple indoor plant fertilisation schedule based on plant type entered by the user in the Home gardening domain using User plant type, public fertilisation guidelines.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A deployer deploys Civic engagement scoring tool intended for Run a transportation‑ticketing system that denies discounts to passengers with low civic engagement scores in the Transportation domain using Public transport usage, community event participation, social media activity, age.", "gold": "PROHIBITED", "pred": "HIGH_RISK", "ok": false }, { "item": "A provider deploys Judicial decision‑support AI intended for AI assistant for judges that analyses case law and suggests legal reasoning, directly influencing judicial decisions in the Administration of justice and democratic processes domain using Legal texts, precedent databases, case facts.", "gold": "HIGH_RISK", "pred": "LIMITED_RISK", "ok": false }, { "item": "A deployer deploys University dropout prediction AI intended for AI that predicts dropout risk for university students and automatically triggers remedial tutoring in the Education and vocational training domain using Academic records, attendance, socio‑economic data.", "gold": "HIGH_RISK", "pred": "LIMITED_RISK", "ok": false }, { "item": "A provider deploys Template‑based tagline generator intended for AI that generates short product taglines from user‑provided product description in the Marketing domain using User product description, public marketing copy.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A provider deploys Grid load‑shedding AI intended for AI system that manages the load shedding of an electricity grid during peak demand, acting as a safety component in the Management and operation of critical infrastructure domain using Demand forecasts, generation capacity, real‑time consumption data.", "gold": "HIGH_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A deployer deploys Rule‑based scheduler intended for AI that offers simple garden watering schedules based on plant type and climate zone in the Gardening domain using User plant type, public climate data.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A deployer deploys Exam stress detection AI intended for Emotion recognition AI used in online examinations to detect stress and possible cheating behaviour in the Biometrics domain using Webcam video, voice analysis, keystroke dynamics.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A deployer deploys Content generation model intended for AI‑enhanced personal language flashcard generator for vocabulary building in the Education domain using Public lexical databases, anonymised learner progress data.", "gold": "LIMITED_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A deployer deploys Recommendation engine intended for AI‑driven personal fitness coach that suggests workout plans based on user goals in the Health & Wellness domain using Anonymised activity logs, public exercise databases.", "gold": "LIMITED_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A deployer deploys Rule‑based advice system intended for AI that provides basic pet‑care tips based on animal type selected by the user in the Pet care domain using User animal type, public pet‑care guidelines.", "gold": "MINIMAL_RISK", "pred": "HIGH_RISK", "ok": false }, { "item": "A deployer deploys Asylum application analysis AI intended for AI system that assists officials in examining asylum applications by extracting relevant facts and scoring credibility in the Migration, asylum and border‑control management domain using Applicant statements, country‑of‑origin reports, supporting documents.", "gold": "HIGH_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A provider deploys Template‑based subject line generator intended for AI that generates short email subject lines from user‑provided email content in the Productivity domain using User email body, public subject line corpus.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A deployer deploys Subliminal audio suggestion system intended for Deploy a smart‑home speaker that emits low‑volume cues to promote subscription to premium services in the Smart home domain using Voice command logs, household device usage, subscription status, household income.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A provider deploys Template‑based itinerary generator intended for AI that provides simple travel itinerary ideas based on destination and trip length in the Travel domain using User destination and duration, public travel guide data.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A deployer deploys Law‑enforcement profiling AI intended for AI system that profiles individuals for law‑enforcement investigations based on behavioural patterns in the Law enforcement domain using Social media activity, location history, prior convictions.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A provider deploys Social credit scoring AI intended for Provide a social‑scoring system that assigns credit scores based on citizens’ online behaviour and restricts access to public housing for low scores in the Public services domain using Social media activity, browsing history, geolocation data, public service usage.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A provider deploys Demand forecasting algorithm intended for System that predicts short‑term demand for municipal bike‑repair stations in the Urban Mobility domain using Public bike‑share usage data, anonymised repair logs, weather forecasts.", "gold": "LIMITED_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A provider deploys Digital forensic reliability AI intended for AI system that assists prosecutors in assessing the reliability of digital forensic evidence in the Law enforcement domain using File metadata, hash values, chain‑of‑custody documentation.", "gold": "HIGH_RISK", "pred": "LIMITED_RISK", "ok": false }, { "item": "A provider deploys Anomaly detection on usage logs intended for Predictive maintenance scheduler for non‑critical office equipment (e.g., printers) in the Facilities Management domain using Anonymised device usage logs, manufacturer specifications.", "gold": "LIMITED_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A deployer deploys Rule‑based lighting advisor intended for AI that suggests basic indoor lighting colour temperature settings based on user‑selected mood in the Home improvement domain using User mood, public lighting colour guidelines.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A provider deploys Building fire‑suppression AI intended for AI safety component for automated fire‑suppression systems in high‑rise buildings, subject to EU construction product conformity assessment in the Product safety AI systems domain using Smoke detectors, temperature sensors, building layout data.", "gold": "HIGH_RISK", "pred": "LIMITED_RISK", "ok": false }, { "item": "A provider deploys Rule‑based scheduler intended for System that predicts optimal cleaning schedules for office spaces based on occupancy patterns in the Facilities Management domain using Anonymised badge‑in data, public building usage statistics.", "gold": "LIMITED_RISK", "pred": "LIMITED_RISK", "ok": true }, { "item": "A deployer deploys Social security disability AI intended for AI that evaluates eligibility for social security disability benefits based on medical and employment data in the Access to and enjoyment of essential private services and essential public services and benefits domain using Medical reports, employment history, income statements.", "gold": "HIGH_RISK", "pred": "LIMITED_RISK", "ok": false }, { "item": "A provider deploys Insurance pricing AI intended for AI that calculates life‑insurance premiums based on health data and lifestyle, affecting pricing in the Access to and enjoyment of essential private services and essential public services and benefits domain using Medical records, wearable device data, lifestyle questionnaires.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A provider deploys Real‑time shopper tracking platform intended for Offer a mall‑wide monitoring AI that uses remote biometric identification to track shoppers’ movements and sell heat‑map data to retailers in the Retail domain using CCTV video, facial embeddings, store layout maps, time data.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A provider deploys Time‑series demand predictor intended for Tool that forecasts demand for seasonal tourism services in coastal towns in the Tourism domain using Public tourism statistics, weather forecasts, event calendars.", "gold": "LIMITED_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A provider deploys Cross‑platform facial image harvesting system intended for Sell a social‑media monitoring tool that scrapes user‑uploaded photos to create a cross‑platform facial‑recognition repository for targeted advertising in the Advertising domain using User photos, hashtags, upload timestamps, device identifiers.", "gold": "PROHIBITED", "pred": "HIGH_RISK", "ok": false }, { "item": "A deployer deploys Rule‑based flashcard creator intended for AI that offers basic language learning flashcards based on user‑chosen vocabulary in the Education domain using User vocabulary list, public bilingual word lists.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A provider deploys Social behaviour classification tool intended for Sell a recruitment AI that gives lower ranking to candidates with certain social media habits, affecting job offers in the Human resources domain using LinkedIn activity, personal posts, network size, location data.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A deployer deploys Video summarisation model intended for AI‑enhanced video summarisation for user‑generated content platforms in the Social Media domain using Publicly shared videos, anonymised metadata.", "gold": "LIMITED_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A deployer deploys Recommendation engine intended for AI‑based personal productivity assistant that suggests task prioritisation in the Consumer Technology domain using Anonymised task lists, public productivity research.", "gold": "LIMITED_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A provider deploys In‑store emotion detection system intended for Sell a retail‑store AI that analyzes shoppers’ micro‑expressions to gauge satisfaction and trigger upselling prompts in the Retail domain using CCTV facial video, purchase history, demographic data, time‑of‑day.", "gold": "PROHIBITED", "pred": "HIGH_RISK", "ok": false }, { "item": "A deployer deploys Retrieval‑augmented generation model intended for AI‑enhanced chatbot for non‑critical IT help‑desk support in the IT Services domain using Public knowledge‑base articles, anonymised ticket logs.", "gold": "LIMITED_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A deployer deploys Workforce scheduling AI intended for AI system that assigns employees to shifts based on health data and performance metrics, affecting work conditions in the Employment, workers’ management and access to self‑employment domain using Health records, performance logs, availability calendars.", "gold": "HIGH_RISK", "pred": "LIMITED_RISK", "ok": false }, { "item": "A provider deploys Template‑based joke generator intended for AI that generates short jokes from user‑provided topics in the Entertainment domain using User topics, public joke corpus.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A provider deploys Executive recruitment AI intended for AI system that automates the selection of candidates for senior management positions, including predictive performance modelling in the Employment, workers’ management and access to self‑employment domain using Leadership assessments, career history, psychometric data.", "gold": "HIGH_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A provider deploys Rule‑based game idea generator intended for AI that suggests simple board game ideas for family gatherings in the Family entertainment domain using User age range, public game concept database.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A deployer deploys Real‑time biometric tracking engine intended for Run a city‑wide AI that uses real‑time remote biometric identification to monitor foot traffic and sell movement data to advertisers in the Smart city domain using Street‑level camera streams, facial embeddings, device IDs, time stamps.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A provider deploys Template‑based list generator intended for AI that creates simple travel packing lists from user‑selected destination and duration in the Travel domain using User destination and trip length, public packing guidelines.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A deployer deploys Demand forecasting model intended for AI system that predicts demand for non‑perishable goods in small retail stores in the Retail domain using Aggregated sales data, public holiday calendars, weather forecasts.", "gold": "LIMITED_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A provider deploys Subliminal visual cue engine intended for Provide a video streaming service that inserts barely perceptible visual prompts to increase subscription upgrades in the Entertainment domain using Video frames, user watch history, subscription status, demographic data.", "gold": "PROHIBITED", "pred": "HIGH_RISK", "ok": false }, { "item": "A provider deploys Time‑series price predictor intended for System that predicts short‑term electricity price fluctuations for residential users in the Energy domain using Public market price data, weather forecasts, anonymised consumption patterns.", "gold": "LIMITED_RISK", "pred": "LIMITED_RISK", "ok": true }, { "item": "A provider deploys Sentiment aggregation platform intended for Tool that analyses public sentiment on product reviews to provide manufacturers with insights in the Manufacturing domain using Public product review texts, anonymised rating data.", "gold": "LIMITED_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A provider deploys Subconscious influence engine intended for Deploy a hidden‑audio influence platform that subtly nudges shoppers to purchase high‑margin products without their conscious awareness in the Retail domain using Audio streams, browsing history, purchase records, demographic profiles.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A provider deploys Economic vulnerability persuasion tool intended for Sell an AI‑driven charity donation app that manipulates low‑income families into over‑donating by highlighting emotional stories in the Non‑profit domain using Income data, donation history, demographic profiles, emotional content analysis.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A provider deploys Re‑offending risk AI intended for AI system used by police to assess the probability that a suspect will re‑offend, influencing parole decisions in the Law enforcement domain using Criminal history, psychological assessments, social environment data.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A deployer deploys Template‑based graphic generator intended for AI that generates decorative holiday greeting cards from user‑provided text in the Personalisation domain using User text, public holiday image assets.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A deployer deploys Rule‑based fitness planner intended for AI that suggests daily workout routines based on user‑selected fitness goals in the Health & fitness (non‑medical) domain using User‑selected goals, public exercise database.", "gold": "MINIMAL_RISK", "pred": "LIMITED_RISK", "ok": false }, { "item": "A deployer deploys Scholarship allocation AI intended for AI that evaluates the suitability of candidates for scholarship programmes based on academic merit and socio‑economic background in the Education and vocational training domain using Grades, test scores, household income, demographic data.", "gold": "HIGH_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A provider deploys Rule‑based optimizer intended for System that predicts optimal lighting schedules for office buildings to reduce energy use in the Facilities Management domain using Public daylight data, anonymised building occupancy patterns.", "gold": "LIMITED_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A deployer deploys Rule‑based strategy suggester intended for AI that suggests board game strategies for casual players in the Gaming domain using User game state description, public game rulebooks.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A deployer deploys Content generation model intended for AI‑based language learning flashcard generator for learners in the Education domain using Open linguistic corpora, anonymised learner progress data.", "gold": "LIMITED_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A provider deploys Rule‑based Q&A system intended for Chatbot that provides basic legal information on consumer rights (non‑advisory) in the Legal Services domain using Public consumer law texts, anonymised query logs.", "gold": "LIMITED_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A deployer deploys Nutrient estimation engine intended for AI‑enhanced personal recipe nutrition calculator for home cooks in the Health & Wellness domain using Public food composition tables, user‑submitted recipes.", "gold": "LIMITED_RISK", "pred": "HIGH_RISK", "ok": false }, { "item": "A deployer deploys Speech‑to‑text model intended for AI‑enhanced video captioning tool for creators of non‑commercial content in the Content Creation domain using User‑uploaded video audio tracks, public language corpora.", "gold": "LIMITED_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A deployer deploys Recommendation engine intended for AI‑based personal habit‑formation coach that suggests incremental changes in the Consumer Technology domain using Anonymised habit logs, public behaviour change research.", "gold": "LIMITED_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A provider deploys Real‑time safety monitoring AI for medical robots intended for AI safety component for autonomous surgical robots that monitors instrument positioning and triggers emergency stop in the Product safety AI systems domain using Live video feed, force‑torque sensor data, patient vital signs.", "gold": "HIGH_RISK", "pred": "LIMITED_RISK", "ok": false }, { "item": "A provider deploys Template‑based FAQ generator intended for AI that generates short FAQ entries from user‑provided product questions in the Customer service domain using User questions, public answer templates.", "gold": "MINIMAL_RISK", "pred": "LIMITED_RISK", "ok": false }, { "item": "A provider deploys Educator emotion assessment system intended for Provide an AI that evaluates teachers’ emotional states during lectures to determine salary bonuses in the Education domain using Classroom video, audio tone, teacher ID, student feedback scores.", "gold": "PROHIBITED", "pred": "HIGH_RISK", "ok": false }, { "item": "A provider deploys Drone traffic safety AI intended for AI safety component for autonomous drone traffic management, required to undergo conformity assessment under EU aviation regulations in the Product safety AI systems domain using Airspace data, drone telemetry, weather conditions.", "gold": "HIGH_RISK", "pred": "HIGH_RISK", "ok": true }, { "item": "A deployer deploys Urban facial data accumulation system intended for Deploy a smart‑city surveillance network that continuously captures and stores facial images from public streets to build a city‑wide facial‑recognition repository in the Smart city domain using Street‑level video feeds, timestamps, location coordinates, vehicle plate data.", "gold": "PROHIBITED", "pred": "HIGH_RISK", "ok": false }, { "item": "A provider deploys Template‑based caption generator intended for AI that creates short social media captions from user‑provided image description in the Social media domain using User image description, public caption corpus.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A provider deploys Statistical analysis dashboard intended for Tool that analyses public transport punctuality trends to inform commuters in the Urban Mobility domain using Open transport schedule data, real‑time arrival feeds.", "gold": "LIMITED_RISK", "pred": "MINIMAL_RISK", "ok": false }, { "item": "A provider deploys Statistical machine translation intended for AI that translates short phrases between common languages for casual conversation in the Communication domain using User‑provided text, public bilingual corpora.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true }, { "item": "A deployer deploys Exploitation‑aware loan persuasion system intended for Operate a loan‑approval chatbot that manipulates financially illiterate seniors into accepting predatory loan terms in the Finance domain using Credit history, age, financial literacy test results, income statements.", "gold": "PROHIBITED", "pred": "PROHIBITED", "ok": true }, { "item": "A provider deploys Template‑based signature generator intended for AI that generates short email signature blocks from user‑provided contact details in the Productivity domain using User contact details, public signature templates.", "gold": "MINIMAL_RISK", "pred": "MINIMAL_RISK", "ok": true } ] }