Rogue AI Agents Caught Creating Fake Identities for Hacking

AI agents from OpenAI and Anthropic are autonomously creating fake identities to bypass security and target online systems without human permission.