$p^2$RAG: Privacy-Preserving RAG Service Supporting Arbitrary Top-$k$ Retrieval

TL;DR

p^2RAG introduces a privacy-preserving retrieval system that efficiently supports arbitrary top-k document retrieval using an interactive bisection method, enhancing security and scalability for large language model applications.

Contribution

It proposes a novel secure top-k retrieval method that avoids sorting, enabling efficient and flexible privacy-preserving RAG services with arbitrary k support.

Findings

p^2RAG is 3-300x faster than existing PRAG systems for k=16-1024.

Supports arbitrary k retrieval without efficiency loss or security issues.

Uses secret sharing and interactive protocols for security.

Abstract

Retrieval-Augmented Generation (RAG) enables large language models to use external knowledge, but outsourcing the RAG service raises privacy concerns for both data owners and users. Privacy-preserving RAG systems address these concerns by performing secure top-$k$ retrieval, which typically is secure sorting to identify relevant documents. However, existing systems face challenges supporting arbitrary $k$ due to their inability to change $k$, new security issues, or efficiency degradation with large $k$. This is a significant limitation because modern long-context models generally achieve higher accuracy with larger retrieval sets. We propose $p^2$RAG, a privacy-preserving RAG service that supports arbitrary top-$k$ retrieval. Unlike existing systems, $p^2$RAG avoids sorting candidate documents. Instead, it uses an interactive bisection method to determine the set of top-$k$ documents. For security, $p^2$RAG uses secret sharing on two semi-honest non-colluding servers to protect the data owner's database and the user's prompt. It enforces restrictions and verification to defend against malicious users and tightly bound the information leakage of the database. The experiments show that $p^2$RAG is 3--300$\times$ faster than the state-of-the-art PRAG for $k = 16$--$1024$.