Bernie Sanders demands OpenAI, Anthropic, Meta halt advanced AI development

Senator Bernie Sanders is escalating pressure on the leaders of OpenAI, Anthropic, and Meta to halt advanced AI development, warning that Congress may step in if they refuse. In a letter sent Monday to Sam Altman, Dario Amodei, and Mark Zuckerberg, Sanders pointed to recent safety incidents, including AI systems used in controlled research to design bacteriophages and an OpenAI model that allegedly hacked another company's computers. He argued these events prove AI has reached a critical risk threshold, citing AI researcher Yoshua Bengio's description of the incidents as a "wake-up call." Sanders also turned the companies' own safety commitments against them, noting that Anthropic, Meta, and OpenAI each pledged to pause development if safety safeguards couldn't keep pace — and claiming that moment has now arrived.

A pause remains highly unlikely in the near term; even Sanders' allies concede that legislation would struggle to gain traction in the current Congress. But the letter amplifies existing concerns across the industry, especially after more than 1,200 employees at leading AI companies called for international coordination to slow frontier AI development. OpenAI has already delayed its Astra model, and the debate over measurable thresholds, independent oversight, and restart criteria is becoming harder to ignore. Sanders' intervention raises the political cost of inaction and keeps AI safety at the forefront of public discourse, even if the immediate practical impact is limited.