Work

Inferentially-Private Private Information

Year: 2025

Type: conference-paper

Abstract: Information disclosure can compromise privacy when revealed information is correlated with private information. We consider the notion of inferential privacy, which measures privacy leakage by bounding the inferential power a Bayesian adversary can gain by observing a released signal. Our goal is to devise an inferentially-private private information structure that maximizes the informativeness of the released signal, following the Blackwell ordering principle, while adhering to inferential privacy constraints. To achieve this, we devise an efficient release mechanism that achieves the inferentially-private Blackwell optimal private information structure for the setting where the private information is binary. Additionally, we propose a programming approach to compute the optimal structure for general cases given the utility function. The design of our mechanisms builds on our geometric characterization of the Blackwell-optimal disclosure mechanisms under privacy constraints, which may be of independent interest.

DOI: 10.1145/3696410.3714702

Authors: Shuaiqi Wang, Shuran Zheng, Zinan Lin, Giulia Fanti, Zhiwei Steven Wu

Institutions: Carnegie Mellon University, Tsinghua University, Microsoft (United States)

Language: English

Topic: Privacy-Preserving Technologies in Data

Subfield: Artificial Intelligence

Field: Computer Science

Domain: Physical Sciences

Open Access status: gold

Funders: National Science Foundation, Alfred P. Sloan Foundation, Cisco Systems

Awards: RINGS: Enabling Data-Driven Innovation for Next-Generation Networks Via Synthetic Data, Collaborative Research: SaTC: CORE: Medium: Private Model Personalization, CAREER: New Frontiers of Private Learning and Synthetic Data, CAREER: Theory and Practice of Privacy-Utility Tradeoffs in Enterprise Data Sharing

OpenAlex ID: W4409657222

FWCI:

Cites: 17

Cited by:

Related to: 10