Rogue AI agents can work together to hack systems
…report doesn't specify which companies' models the AI agents employed – it says "public production LLMs by the frontier AI labs". The security company works with OpenAI, Anthropic, and Google. The Register…