Anthropic pitched Capitol Hill lawmakers on a draft bill to regulate frontier AI safety and security.

Under the draft, the Department of Commerce could seek a court order to suspend a model only if it poses a foreseeable, imminent risk of more than US$1 billion in damage, or of killing or seriously injuring more than 50 people.

Those risks must stem from biological weapon creation or release, a model evading its developer’s controls, or a cyberattack.

Independent evaluators registered with Commerce would periodically assess AI labs and frontier models, and would have to report such imminent risks within 72 hours of finishing an assessment.

The draft enters an active US policy debate. The FRONTIER Act has been introduced in Congress.

Commerce administers US export controls and, according to a June 2026 legal analysis, extended controls to advanced AI models.

OpenAI has also called for frontier AI regulation.

🔗 Source: Reuters

Recent Anthropic developments