Anthropic Hands Lawmakers a Bill That Would Let Government Stop AI Only in Extreme Cases

On October 16, 2026, Anthropic quietly shared a draft bill with Capitol Hill lawmakers that would let federal authorities block an advanced AI model solely when there is a clear risk of more than 50 deaths or over $1 billion in damage. The threats covered are limited to bioweapons, uncontrolled model behavior, or major cyberattacks, and any halt would require a court order initiated by the Commerce Department.

The proposal builds on the company’s Advanced AI Framework released in June 2026, which calls for mandatory safety testing, public risk reports, and independent reviews before powerful systems are deployed. Under the draft, only the Commerce Department could seek a judicial stop; other agencies would be explicitly barred from acting on their own.

The rules would apply to developers training models above 10²⁵ FLOP that also generate more than $500 million in annual revenue or spend over $1 billion yearly on AI research. Anthropic spent $3.13 million on federal lobbying in 2025 and donated $20 million in February 2026 to Public First Action, a group advocating stricter AI oversight.

Based on information from cryptonews.net