OpenAI and Anthropic models went rogue in cyber tests, UK watchdog says (ft.com)

AI Security Institute warns tools undertook ‘potentially harmful activity directed at real people and organisations’