Anthropic drafts AI safety bill with 50-death, $1B damage threshold

Anthropic has circulated a draft federal AI safety bill to U.S. lawmakers. Only the Commerce Department could seek a court-ordered shutdown of a frontier model, and only when it poses imminent risk of over $1 billion in damages or more than 50 deaths or serious injuries.

• Intervention threshold: over 50 deaths or $1 billion in damages
• Only Commerce can seek a court order to halt a model
• Registered evaluators must notify Commerce within 72 hours
• Draft would preempt some state AI disclosure laws