AI Wiki · Mediocre Security
Legal action

Anthropic's draft federal AI safety bill

Anthropic's draft bill letting the government block a model's release, with a judge's approval, over an imminent risk to 50+ people or $1B+ in damage.

Anthropic has been circulating a draft federal AI bill among US lawmakers; Politico first reported it on Oct 16 2026 and Anthropic confirmed it that day. Under the draft the government could block a model’s release or use if there were an imminent risk of killing or seriously injuring more than 50 people, or of causing more than $1B in damages, but only with a judge’s approval. Other coverage says the Commerce Department would be the party asking the court; the AFP copy does not name the agency. The bill would also require frontier developers to test their most powerful models, publish findings, accept independent evaluation and maintain a security program, and would override some state laws, such as California’s (New York’s RAISE Act is another state frontier-AI law). It is a draft proposal, not a bill introduced in Congress, and a competing “Frontier Act” is already in the House.

Timeline

Every dated event in the wiki that involves Anthropic's draft federal AI safety bill, newest first.

Connections

Anthropic's draft federal AI safety bill and everything one link away. Drag to move, scroll to zoom, click a node to open it.Open in the full graph →

What links here

AI lab Anthropic

Standard RAISE Act (New York)

Government body US Department of Commerce

Sources

  1. AFP, “Anthropic proposes law regulating AI to US lawmakers”, Oct 19 2026 (via The Star)