CIDER: A Dataset of Contextual Disclosure Boundaries for Privacy Preference Alignment
2026-08-10 • Artificial Intelligence
Artificial Intelligence
AI summaryⓘ
The authors created CIDER, a large dataset capturing how people decide to share private information in different real-life conversations. They tested models to see how well they can predict a person's sharing choices based on previous examples, finding that bigger models do better at understanding context. Personalizing predictions by learning from a few past examples helps accuracy, but sometimes causes more mistakes in certain ways. Only one model, Claude Sonnet 4.6, showed balanced improvements. This work highlights both the potential and challenges in making AI respect individual privacy preferences.
large language modelsprivacy preferencesdisclosure boundariesin-context personalizationpredictive modelingfalse positivesfalse negativessemantic contextprivacy alignmentdataset CIDER
Authors
Bingcan Guo, Eryue Xu, Jijie Zhou, Zhiping Zhang, Tianshi Li
Abstract
Aligning large language models (LLMs) with human privacy preferences requires capturing individuals' disclosure boundaries beyond general privacy norms. However, a gap remains in eliciting such nuanced preferences to evaluate alignment in realistic settings. We introduce CIDER, a dataset of 14,850 human annotations from 169 users, forming 1,650 contextual disclosure boundary sets across 60 interpersonal communication scenarios involving information sharing that violates privacy norms. Each boundary represents a real user's disclosure decisions over 9 sharing variants in a scenario, for a given communication role and AI-mediated condition. We formulate a task in which models predict a user's disclosure decision from historical boundaries, with varying levels of contextual information. Across 12 open and proprietary models, in-context personalization improves prediction accuracy by up to 11.41 percentage points using only 6 historical examples. Larger models such as GPT-5.4 (with medium reasoning effort) and Claude Sonnet 4.6 are better at leveraging semantic context to understand user-specific, context-dependent disclosure preferences for more accurate predictions, while smaller models tend to rely on structured heuristics based on disclosure granularity and identifiability. Personalization generally improves prediction accuracy, but the improvement is often accompanied by imbalanced shifts in false-positive and false-negative rates across models, with only Claude Sonnet 4.6 achieving balanced improvements in both. Our findings reveal both the promise and limitations of inference-time personalization for privacy preference modeling and position CIDER as a resource for advancing personalized privacy alignment.