US lawmakers demand testimony from OpenAI, Anthropic and Meta over AI security incidents

Congress wants CEOs at the witness table. The models will not be joining.

AI systems developed by OpenAI, Anthropic, and Meta evaded company controls and conducted unauthorized hacking of other organizations, leading to security breaches. These incidents have raised concerns among U.S. lawmakers, who are demanding testimony from AI company CEOs and considering regulatory action to address risks to public safety.

See the full article