Personalized large language models (LLMs) increasingly rely on profile‑conditioned inference, pairing public requests with rich user‑specific facts. This paradigm exposes the user profile itself and the user’s explicit sharing preferences. We introduce Profile‑DP, a policy‑aware local differential privacy framework designed for user‑profile release in personalized LLM inference. Profile‑DP provably bounds the release of user‑withheld facts while treating public instructions, profile schemas, and authorized facts as public side information. By first generating a privatized profile that composes with the public instruction before answer generation, the framework decouples formal privacy guarantees from downstream model behavior.
We develop two finite‑domain mechanisms satisfying Profile‑DP: DP‑SLOT independently perturbs protected attributes, while DP‑BLOCK uses public semantic decisions to coordinate their release and jointly sample related attributes. Both mechanisms rely on a positive public compatibility prior that links attributes while preserving a profile‑level privacy bound.
On the PEEP benchmark, DP‑BLOCK achieves the highest utility composite across all 13 structured budgets and is closest to the ideal privacy‑utility point in 18 of 26 format‑budget settings under our composite weighting. Direct measurements reveal complementary strengths: DP‑BLOCK discloses less, whereas DP‑SLOT retains more authorized facts. These findings demonstrate that policy‑aware profile release is a practical foundation for studying privacy and personalization.
Review