SkillCloak Lets Malicious AI Agent Skills Evade Static Scanners with Self-Extracting Packing

Researchers from Hong Kong University of Science and Technology have unveiled a new technique called SkillCloak that allows malicious AI agent skills to evade static scanners, fooling them over 90% of the time. This self-extracting packing method keeps the harmful code active while avoiding detection. The implications are significant, as it raises concerns over the effectiveness of current security measures against advanced AI-based threats. This development could prompt a reevaluation of how we protect AI systems from malicious code.