Generating Privacy-Preserving Personalized Advice with Zero-Knowledge Proofs and LLMs

TL;DR

This paper presents a novel framework combining zero-knowledge proofs with large language models to enable privacy-preserving personalized advice, addressing privacy concerns in sensitive data domains.

Contribution

It introduces an architecture and prompting strategy that integrate zkVM with LLMs, demonstrating practical feasibility and highlighting current limitations.

Findings

Empirical evaluation shows the approach is feasible in real-world scenarios.

Identifies current constraints and performance limitations of zkVM and prompting strategies.

Provides a new method for privacy-preserving personalization in AI systems.

Abstract

Large language models (LLMs) are increasingly utilized in domains such as finance, healthcare, and interpersonal relationships to provide advice tailored to user traits and contexts. However, this personalization often relies on sensitive data, raising critical privacy concerns and necessitating data minimization. To address these challenges, we propose a framework that integrates zero-knowledge proof (ZKP) technology, specifically zkVM, with LLM-based chatbots. This integration enables privacy-preserving data sharing by verifying user traits without disclosing sensitive information. Our research introduces both an architecture and a prompting strategy for this approach. Through empirical evaluation, we clarify the current constraints and performance limitations of both zkVM and the proposed prompting strategy, thereby demonstrating their practical feasibility in real-world scenarios.