This research introduces a framework integrating zero-knowledge proofs with LLM-based chatbots to deliver personalized advice while protecting user privacy.

• Enables privacy-preserving personalization for sensitive domains like finance, healthcare, and relationships
• Implements zkVM technology to verify user attributes without exposing raw data
• Creates a system where users can receive tailored advice while maintaining data minimization
• Addresses critical security concerns in AI personalization by keeping sensitive information private

This innovation represents a significant advancement for the security field by demonstrating how AI systems can provide personalized services without compromising user privacy—essential as LLMs become more integrated into sensitive domains.

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