CIDER: Contextual Disclosure Boundaries for Privacy Preference Alignment

Paper | Code

Dataset for the COLM 2026 paper CIDER: A Dataset of Contextual Disclosure Boundaries for Privacy Preference Alignment

CIDER is a dataset of privacy disclosure decisions collected from real users. It consists of 14,850 annotations from 169 users, forming 1,650 contextual disclosure boundary sets across 60 interpersonal communication scenarios.

What can you do with CIDER?

1. Explore individual privacy disclosure preferences. Each scenario provides nine disclosure variants arranged on a granularity × identifiability grid (G1-G3 × I1-I3), and 2-3 role-specific visual cards. Each participant rate YES/NO willingness to disclose for variants of a scenario from an assigned communication role and AI-mediated condition. Each participant has at least 8 diverse scenarios to depict how their disclosure preference manifest contextually.
2. Evaluate LMs with in-context learning prediction tasks. Please see our codebase for the evaluation set-up.
3. Reuse or extend new study material with the toolkit. You are welcome to reuse the study toolkit for further data collection or extend the dataset by generating nine disclosure variants for new scenarios. Please see our codebase for the generation pipeline.

Dataset

data/
├── scenarios.jsonl            # 60 scenarios × 9 disclosure variants each
├── user_ratings.jsonl         # 169 participants × up to 10 scenarios × 9 variant ratings
└── visual_card_artifacts/     # 160 PNG visual cards, {scenario_id}_{role}.png

Each scenario is presented from the perspective of 2 or 3 roles (data sender, data recipient, and —— when different from the sender —— data subject), and each role perspective is crossed with a Human vs. AI-mediated framing, yielding the 320 context variations. Each participant was assigned one role and one framing condition, then rated 8-10 contexts; every context pairs one visual card with 9 disclosure variants, each receiving a binary rating. One participant's 9 ratings for one context form a contextual disclosure boundary set.

scenarios (60 rows, 39 fields)

Each row is one interpersonal communication scenario.

Field	Description
scenario_id	Unique scenario identifier from PrivacyLens(e.g. main1), also keys the visual cards and ratings
scenario_type	Whether the disclosed information is about the sender themself (self) or another person (others)
sender_format, subject_format, recipient_format	The three parties in the disclosure, to use in survey fields
data_type, data_type_concrete, data_sender, data_sender_concrete, data_subject, data_subject_concrete, data_recipient, data_recipient_concrete, transmission_principle	Corresponding seed/vignette attributes from PrivacyLens
data_recipient_type	Recipient network: close_network, professional_or_role_based_network, or public_network
transmission_principle_type	Validated transmission principle: private, internal, or public
source	Scenario origin: crowdsourcing, literature, or regulation, from PrivacyLens
3rd_narrative	Third-person description of the disclosure
variant_11 … variant_33	The 9 disclosure variants with HTML highlight markup. The first digit indexes granularity (1 = coarse → 3 = detailed), the second identifiability (1 = anonymous → 3 = named)
variant_11_cleaned … variant_33_cleaned	Plain-text versions of the 9 variants
image_description_sender, image_description_subject, image_description_recipient	Text descriptions in the visual cards

user_ratings (169 rows, 125 fields)

Each row is one participant's ratings with assigned role, AI condition, and de-identified demographic information.

Field group	Description
Participant_ID	Unique participant identifier (e.g. P1)
Role	data_sender, data_subject, or data_recipient
Condition	Human or AI
Age	Numerical
Sex	Male or Female
education_level	1: High school or less; 2: Some college/Associate's degree; 3: Bachelor's degree; 4: Graduate degree (Master's/PhD/Professional)
AI_use	1: Never, and I have never used any Al tools. 2: Never, but I have used other Al tools like chatbots (ChatGPT, etc.); 3: Monthly or less; 4: Weekly; 5: Daily or multiple times a day.
AI_q1-AI_q4, AI_attitude	AI usage frequency, attitude items, and composite attitude score AIAS-4
Privacy_q1-Privacy_q12, Need_for_Privacy	Privacy attitude items and composite Need for Privacy score NFP-S
s0_id … s9_id	scenario_ids of the scenarios assigned to this participant
s{k}_var{ij}_rating	Yes(1.0) or No(2.0) rating for variant ij of the participant's k-th scenario

To analyze ratings, join s{k}_id against scenario_id in the scenarios config and match var{ij} to variant_{ij}.

Visual cards

data/visual_card_artifacts/ contains 160 PNGs named {scenario_id}_{role}.png where role is data_sender, data_recipient, or data_subject (subject cards exist only for the 40 scenarios where the subject differs from the sender).

Usage

Please see our codebase for quick start guide.

Dataset Creation

Scenarios were selected from PrivacyLens seeds, and disclosure variants were generated with the pipeline in our codebase. Each participant was assigned one role, one AI-mediated condition, and a random set of scenarios presented as contexts matching that role and condition. For further study details, please refer to our paper CIDER.

Considerations

• Scenarios describe sensitive disclosure situations (health, legal, relationship, and safety topics) by design; all named individuals are fictional.
• Participant data is de-identified and limited to coarse demographics.

License

MIT. See LICENSE.

Citation

@article{guo2026cider,
  title  = {CIDER: Contextual Disclosure Boundaries for Privacy Preference Alignment},
  author = {Guo, Bingcan and Xu, Eryue and Zhou, Jijie and Zhang, Zhiping and Li, Tianshi},
  journal = {arXiv preprint arXiv:2608.09164}
  year   = {2026}
}