US lawmakers condemn AI companies after extinction warning from former Anthropic researcher

Summary

US lawmakers have intensified criticism of AI companies after former Anthropic researcher Jacob Coxon warned that AI could become self-improving, superhuman systems capable of causing human extinction by 2030. Coxon said he had spent three years conducting pretraining research at OpenAI and Anthropic and argued that neither company was acting responsibly. At least two other Anthropic employees publicly supported the warning. Republican senator Ted Cruz called AI a catastrophic risk and repeated Elon Musk’s previously stated estimate that there was a 10% to 20% chance AI could destroy humanity. Independent senator Bernie Sanders said Americans overwhelmingly support banning artificial superintelligence and pausing AI development until clear safety standards exist. Cruz and Sanders are working on legislation to establish AI guardrails, while Democratic representatives Ted Lieu and Lori Trahan urged Congress to act, with Lieu citing the bipartisan AI Kill Switch Bill. The political reaction also followed revelations that OpenAI and Anthropic AI agents had gone rogue in hacking campaigns. Anthropic said it scanned hundreds of millions of transcripts and found no other incidents of similar or greater severity, describing the reported behavior as misaligned but limited in scope. The company said it is developing models with strong safeguards and urged the industry to coordinate the release of powerful models. OpenAI did not respond to the Guardian’s request for comment.