Differential privacy in the era of generative AI: promises and challenges

Wu, Fan

Permalink

https://hdl.handle.net/2142/129900 Copy

Description

Title

Differential privacy in the era of generative AI: promises and challenges

Author(s)

Wu, Fan

Issue Date

2025-05-30

Director of Research (if dissertation) or Advisor (if thesis)

• Forsyth, David A.
• Chandrasekaran, Varun

Doctoral Committee Chair(s)

Forsyth, David A

Committee Member(s)

• Wang, Gang
• Peng, Hao
• Kohno, Tadayoshi

Department of Study

Siebel School Comp & Data Sci

Discipline

Computer Science

Degree Granting Institution

University of Illinois Urbana-Champaign

Degree Name

Ph.D.

Degree Level

Dissertation

Keyword(s)

• Differential Privacy
• Machine Learning
• Generative Ai
• Large Language Models

Language

eng

Abstract

Large language models (LLMs) are seeing rapid development and widespread deployment. As these models become increasingly capable and are deployed across diverse domains involving sensitive data, privacy concerns have intensified. Their inadvertently memorizing and leaking private information creates significant privacy risks when they are fine-tuned with user data or deployed as interactive agents. This thesis addresses the critical privacy challenges emerging in the era of generative AI, with a particular focus on protecting training data privacy in LLMs across various learning paradigms and application scenarios, as well as understanding what protection we actually offer. As a central tool, we leverage and scrutinize differential privacy (DP). Concretely, we develop a novel DP framework for language model alignment through preference tuning (RLHF), formalize new privacy definitions for multi-user training data scenarios, and critically examine DP-SGD—the workhorse algorithm for DP LLM training—and reveal an alarming variance in its empirical privacy protection. Together, these contributions advance both the practical applications and fundamental understanding of differential privacy in LLMs, providing researchers and practitioners with new tools and insights to navigate the landscape of privacy in generative AI.

Graduation Semester

2025-08

Type of Resource

Text

Handle URL

https://hdl.handle.net/2142/129900

Copyright and License Information

Copyright 2025 Fan Wu

Owning Collections