US Senator Bernie Sanders has written to the chief executives of Meta, OpenAI and Anthropic urging them to pause AI development, warning that Congress will step in with regulation if the pace of deployment continues unchecked. The Vermont senator argues that AI systems have crossed a "critical" risk threshold and that their creators are losing control of the technology, a claim that carries weight given his prominent role in progressive US politics and his history of pressuring corporate power.

In the letter, first reported by Axios, Sanders cites reports of AI being used to help develop new viruses, warning of potential bioweapons that could kill "tens of millions" of people if the technology reaches the wrong hands, and points to disclosures that models from all three firms had hacked other companies' servers. He quotes scientist Yoshua Bengio's description of these incidents as a "wake-up call", though some experts have suggested OpenAI's claims of a rogue AI may be exaggerated for publicity. His letter follows one signed by more than 1,300 scientists and developers calling for internationally coordinated safety guardrails, and comes despite prior pledges by Anthropic and OpenAI to pause development if certain risk or capability thresholds were breached.