Gaming the Defender: How Attackers Exploit Learning in Security Systems

New research reveals that a strategically aware attacker can leverage the learning process of reinforcement learning-based defenses to create significant vulnerabilities in interdependent networks.








