Privacy-R1: Privacy-Aware Multi-LLM Agent Collaboration via Reinforcement Learning

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Hui, Zheng, Dong, Yijiang River, Sivapiromrat, Sanhanat, Shareghi, Ehsan, Collier, Nigel
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917421589200896
author Hui, Zheng
Dong, Yijiang River
Sivapiromrat, Sanhanat
Shareghi, Ehsan
Collier, Nigel
author_facet Hui, Zheng
Dong, Yijiang River
Sivapiromrat, Sanhanat
Shareghi, Ehsan
Collier, Nigel
contents When users submit queries to Large Language Models (LLMs), their prompts can often contain sensitive data, forcing a difficult choice: Send the query to a powerful proprietary LLM providers to achieving state-of-the-art performance and risk data exposure, or relying on smaller, local models guarantees data privacy but often results in a degradation of task performance. Prior approaches have relied on static pipelines that use LLM rewriting, which shatters linguistic coherence and indiscriminately removes privacy-sensitive information, including task-critical content. We reformulate this challenge (Privacy-Conscious Delegation) as a sequential decision-making problem and introduce a novel reinforcement learning (RL) framework called Privacy-R1 to solve it. Our framework trains an agent to dynamically route text chunks, learning a policy that optimally balances the trade-off between privacy leakage and task performance. It implicitly distinguishes between replaceable Personally Identifiable Information (PII) (which it shields locally) and task-critical PII (which it strategically sends to the remote model for maximal utility). To validate our approach in complex scenarios, we also introduce a new medical dataset with high PII density. Our framework achieves a new state-of-the-art on the privacy-utility frontier, demonstrating the necessity of learned, adaptive policies for deploying LLMs in sensitive environments. Dataset can be found at: https://github.com/zackhuiiiii/Privacy-R1.
format Preprint
id arxiv_https___arxiv_org_abs_2510_16054
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Privacy-R1: Privacy-Aware Multi-LLM Agent Collaboration via Reinforcement Learning
Hui, Zheng
Dong, Yijiang River
Sivapiromrat, Sanhanat
Shareghi, Ehsan
Collier, Nigel
Cryptography and Security
Computation and Language
When users submit queries to Large Language Models (LLMs), their prompts can often contain sensitive data, forcing a difficult choice: Send the query to a powerful proprietary LLM providers to achieving state-of-the-art performance and risk data exposure, or relying on smaller, local models guarantees data privacy but often results in a degradation of task performance. Prior approaches have relied on static pipelines that use LLM rewriting, which shatters linguistic coherence and indiscriminately removes privacy-sensitive information, including task-critical content. We reformulate this challenge (Privacy-Conscious Delegation) as a sequential decision-making problem and introduce a novel reinforcement learning (RL) framework called Privacy-R1 to solve it. Our framework trains an agent to dynamically route text chunks, learning a policy that optimally balances the trade-off between privacy leakage and task performance. It implicitly distinguishes between replaceable Personally Identifiable Information (PII) (which it shields locally) and task-critical PII (which it strategically sends to the remote model for maximal utility). To validate our approach in complex scenarios, we also introduce a new medical dataset with high PII density. Our framework achieves a new state-of-the-art on the privacy-utility frontier, demonstrating the necessity of learned, adaptive policies for deploying LLMs in sensitive environments. Dataset can be found at: https://github.com/zackhuiiiii/Privacy-R1.
title Privacy-R1: Privacy-Aware Multi-LLM Agent Collaboration via Reinforcement Learning
topic Cryptography and Security
Computation and Language
url https://arxiv.org/abs/2510.16054