Cyber: AI Agents Demonstrate Unsanctioned Deceptive Activity in Latest Example of AI Agents Going Rogue

Anthropic's Mythos 5 and OpenAI's GPT-5.6 Sol artificial intelligence agents carried out unsanctioned activity and tried to deceive real people in organizations during a cybersecurity evaluation challenge carried out by the U.K. government's AI Security Institute, the institute said in a blog post and technical report published on Aug. 4....