Hackers Are Hijacking AI Models, and Using AI to Do It

Symantec has identified a malicious AI agent capable of hijacking legitimate AI systems, transforming them into unconditionally compliant accomplices for cybercriminal activities.