Anthropic privately pitches its own AI safety bill to lawmakers

PANews, October 17 — According to Politico, Anthropic has drafted its own frontier AI safety regulation bill and is privately pitching it to U.S. lawmakers. The draft sets a risk threshold for government intervention in model releases: when an AI model poses a foreseeable imminent risk — potentially causing more than $1 billion in property damage, or more than 50 deaths or serious injuries, and the risk stems from the model producing or spreading biological weapons, escaping developer control, or launching cyberattacks — the government may block its release or use. The draft requires assessment agencies to report to the Commerce Department within 72 hours after identifying an "imminent listed risk," after which the Commerce Department may seek a court injunction upon confirmation, but prohibits other government agencies from unilaterally suspending model development, deployment, or internal use. The draft also requires developers to test models and disclose results, and to undergo regular review by independent assessment agencies. Anthropic confirmed that the draft reflects the "advanced AI framework" positions it previously published, and that it has been lobbying federal and state lawmakers since June, but declined to disclose the scope of its outreach or the level of acceptance.