PrivAct: Internalizing Contextual Privacy Preservation via Multi-Agent Preference Training

Abstract

Large language model (LLM) agents are increasingly deployed in personalized tasks involving sensitive, context-dependent information, where privacy violations may arise in agents' action due to the implicitness of contextual privacy. Existing approaches rely on *external*, inference-time interventions which are brittle, scenario-specific, and may expand the privacy attack surface. We propose **PrivAct**, a contextual privacy-aware multi-agent learning framework that *internalizes* contextual privacy preservation directly into models' generation behavior for privacy-compliant agentic actions. By embedding privacy preferences into each agent, PrivAct enhances system-wide contextual integrity while achieving a more favorable privacy-helpfulness tradeoff. Experiments across multiple LLM backbones and benchmarks demonstrate consistent improvements in contextual privacy preservation, reducing leakage rates by up to 12.32\% while maintaining comparable helpfulness, as well as zero-shot generalization and robustness across diverse multi-agent topologies. The code and datasets will be released at [URL/upon acceptance].