Observing concession trajectories and timing allows adversaries to infer private constraints in autonomous negotiation agents. Apple researchers formalized these behavioral privacy leaks and proposed randomized policies to mitigate the risk. This prevents attackers from reverse-engineering sensitive data through observable dynamics. Practitioners must now prioritize behavioral obfuscation alongside standard cryptographic protections.