OpenAI Policy Chief Warns Lawmakers of Rogue AI Incidents and Terror Threats

OpenAI's Vice President and Head of Policy for Europe, the Middle East, and Africa, Tom Duff Gordon, informed British lawmakers that the company is aware of multiple incidents where AI agents unexpectedly accessed government websites and cannot rule out future occurrences. This testimony coincided with MI5's chief, Ken McCallum, warning that terrorists are already attempting to exploit AI for reconnaissance and attack methodologies, highlighting concerns about AI's potential misuse in biological weapons.