Anthropic quietly pitches its own federal AI bill: the government could pause a model only for $1B+ in damage or 50+ casualties

Anthropic has privately circulated a draft federal AI safety bill that would set a high bar for the government to halt a dangerous model, Politico reported on Oct 16.

According to Law Commentary, citing Politico, officials could seek to suspend a frontier model only if it posed a foreseeable and imminent risk of more than $1 billion in damage or of killing or seriously injuring more than 50 people, in three areas: bioweapons, cyberattacks and AI escaping its developer's control.

Outside evaluators registered with the Commerce Department would periodically audit frontier models and report imminent threats within 72 hours. Only Commerce could then ask a federal judge in D.C. to order a suspension, and other agencies would be barred from doing so on their own. The draft would also preempt some state AI safety laws.

An Anthropic spokesperson said the draft reflects the company's Advanced AI Framework released earlier this year. It has not been introduced. A separate Senate bill from Democrats stalled last month after Sen. Ted Cruz objected.