Behavioral Privacy Leakage in Agentic Negotiation: Formalizing and Mitigating Inference Attacks via Randomized Policies
Autonomous agents have a privacy problem nobody's talking about: adversaries can reverse-engineer your negotiation strategy from how you move.
Why it matters
As AI agents take on high-stakes negotiation roles in insurance and procurement, a new class of inference attacks—behavioral privacy leakage—exposes hidden constraints through negotiation dynamics alone. This research formalizes the threat and proposes defenses via randomized policies, highlighting a critical gap between cryptographic protection and real-world agent deployment security.
The key facts
6 to knowAccepted at AI4TCI Workshop (ARES 2026)
Focus on autonomous negotiation agents in high-stakes domains (insurance, procurement)
Threat vector: behavioral privacy leakage via concession trajectories, timing patterns
Defense mechanism: randomized policies
Cryptographic techniques alone insufficient for agent privacy
Published by Apple ML Research
Go to the source
Apple Machine Learningmachinelearning.apple.com
Publisher excerpt: This paper was accepted at the AI4TCI (Workshop on AI for Secure and Trustworthy Critical Infrastructure Systems) Workshop at the International Conference on Availability, Reliability and Security (ARES) 2026. Autonomous negotiation agents are increasingly deployed in high-stakes settings such as…