gdpr-normative-triples / rules /dpia_guidelines_wp248.json
Mujtaba Hussain
Publish 936 GDPR Normative Triples and Knowledge Graph
da77f7a
Raw History Blame Contribute Delete
6.79 kB
{
"title": "Article 35 DPIA Statutory Assessment Framework (EDPB Guidelines WP 248 rev.01)",
"version": "1.0.0",
"celex": "32016R0679",
"legal_mandate": "Regulation (EU) 2016/679 Article 35(1) & Article 35(3)",
"high_risk_criteria": [
{
"id": "CRIT-01",
"name": "Evaluation or Scoring",
"description": "Profiling and predicting aspects concerning the data subject's performance at work, economic situation, health, personal preferences, or behavior.",
"legal_basis": "Art. 35(3)(a), Recital 71, Recital 91",
"example": "Credit scoring, behavioural advertising, employee productivity tracking."
},
{
"id": "CRIT-02",
"name": "Automated Decision-Making with Legal / Significant Effect",
"description": "Processing that aims at taking decisions on data subjects producing legal effects or significantly affecting them.",
"legal_basis": "Art. 22(1), Art. 35(3)(a)",
"example": "Automated mortgage rejection, algorithmic CV screening without human review."
},
{
"id": "CRIT-03",
"name": "Systematic Monitoring",
"description": "Processing used to observe, monitor, or control data subjects, including data collected via public networks or public spaces.",
"legal_basis": "Art. 35(3)(c), Recital 91",
"example": "Public CCTV, vehicle telematics, employee keystroke monitoring, web session replay."
},
{
"id": "CRIT-04",
"name": "Sensitive Data or Data of Highly Personal Nature",
"description": "Special categories of data (Art. 9), criminal convictions (Art. 10), electronic communications data, financial data, or location tracking.",
"legal_basis": "Art. 9(1), Art. 10, Art. 35(3)(b)",
"example": "Health records, biometric vectors, political opinions, union membership."
},
{
"id": "CRIT-05",
"name": "Data Processed on a Large Scale",
"description": "High volume of data, broad geographical extent, large number of data subjects, or permanent duration.",
"legal_basis": "Art. 35(3)(b), Recital 91",
"example": "Nationwide customer loyalty card, telecom location data, national health portal."
},
{
"id": "CRIT-06",
"name": "Matching or Combining Datasets",
"description": "Combining data originating from two or more operations performed for different purposes and/or by different controllers exceeding reasonable expectations.",
"legal_basis": "Recital 91",
"example": "Cross-referencing advertising tracking with offline government census data."
},
{
"id": "CRIT-07",
"name": "Vulnerable Data Subjects",
"description": "Processing data concerning subjects who may not be able to easily consent or oppose processing due to an imbalance of power.",
"legal_basis": "Recital 75, Recital 91",
"example": "Children, patients, employees, elderly individuals, asylum seekers."
},
{
"id": "CRIT-08",
"name": "Innovative Use or Applying New Technologies",
"description": "Use of novel technology or novel organizational solutions that may involve novel forms of data collection and risk.",
"legal_basis": "Art. 35(1), Recital 89, Recital 91",
"example": "Generative AI, facial recognition cameras, Internet of Things (IoT) sensors, neuro-tech."
},
{
"id": "CRIT-09",
"name": "Cross-Border Transfers Outside the EU/EEA",
"description": "Transfers of personal data to a third country or international organization not subject to an adequacy decision.",
"legal_basis": "Chapter V (Art. 44-49)",
"example": "Syncing EU customer telemetry to US or non-EEA cloud clusters without verified SCCs."
},
{
"id": "CRIT-10",
"name": "Processing Preventing Exercise of Rights or Service Use",
"description": "Processing where data subjects cannot use a service or enter into a contract due to algorithmic barriers.",
"legal_basis": "Recital 91",
"example": "Credit reference agency blacklist preventing creation of bank accounts."
}
],
"mitigating_toms": [
{
"id": "TOM-01",
"name": "State-of-the-Art AES-256 KMS Encryption & TLS 1.3",
"category": "TECHNICAL",
"risk_reduction_pct": 25,
"legal_basis": "Art. 32(1)(a)"
},
{
"id": "TOM-02",
"name": "Cryptographic Pseudonymization & Tokenization",
"category": "TECHNICAL",
"risk_reduction_pct": 25,
"legal_basis": "Art. 32(1)(a), Art. 25"
},
{
"id": "TOM-03",
"name": "Certified Human-in-the-Loop Override Mechanism",
"category": "ORGANIZATIONAL",
"risk_reduction_pct": 20,
"legal_basis": "Art. 22(3); AI Act Art. 14"
},
{
"id": "TOM-04",
"name": "Differential Privacy & Training Data Minimization",
"category": "TECHNICAL",
"risk_reduction_pct": 20,
"legal_basis": "Art. 5(1)(c); AI Act Art. 10"
},
{
"id": "TOM-05",
"name": "Strict Role-Based Access Control (RBAC) & Hardware MFA",
"category": "ORGANIZATIONAL",
"risk_reduction_pct": 15,
"legal_basis": "Art. 32(1)(b); NIS2 Art. 21"
},
{
"id": "TOM-06",
"name": "Automated Hard Expiration TTL Purge",
"category": "TECHNICAL",
"risk_reduction_pct": 15,
"legal_basis": "Art. 5(1)(e)"
}
],
"sample_processing_operations": {
"op-ai-recruitment": {
"title": "Autonomous AI Recruitment & Candidate Evaluation Engine",
"description": "Algorithmic screening of applicant video interviews, vocal inflection, and CV history to generate automated suitability scores and pre-screen rejections.",
"active_criteria": ["CRIT-01", "CRIT-02", "CRIT-04", "CRIT-07", "CRIT-08"],
"inherent_likelihood": 4,
"inherent_severity": 4,
"recommended_toms": ["TOM-03", "TOM-02", "TOM-04"]
},
"op-smart-cctv": {
"title": "Commercial Center CCTV Facial Recognition & Demographic Heatmap",
"description": "Continuous real-time video surveillance in retail malls analyzing shopper paths, age, and emotional sentiment to optimize store layouts.",
"active_criteria": ["CRIT-03", "CRIT-04", "CRIT-05", "CRIT-07", "CRIT-08"],
"inherent_likelihood": 5,
"inherent_severity": 5,
"recommended_toms": ["TOM-01", "TOM-02", "TOM-06"]
},
"op-internal-payroll": {
"title": "Internal Employee Monthly Salary & Tax Remittance",
"description": "Processing standard employee bank details and tax codes solely for monthly salary payout and mandatory social security filing.",
"active_criteria": ["CRIT-07"],
"inherent_likelihood": 1,
"inherent_severity": 2,
"recommended_toms": ["TOM-01", "TOM-05"]
}
}
}