Anthropic pitches bill limiting AI blocks to $1bn-damage risks

What happened: Anthropic has privately presented US lawmakers with a draft bill on halting dangerous AI models, according to TechSpot. The report says the story originated with Politico. It is a proposal, not law.

The numbers: The trigger is a foreseeable, imminent risk of more than $1 billion in damage, or death or serious injury to more than 50 people.

The details: Only three risk types would qualify: biological weapons, models escaping their developers' control, and cyberattacks.

How it would work: The Commerce Department would have to make its case in federal court. A judge in the US District Court for DC would decide whether to stop a release.

The context: One report says the draft would expressly prohibit any federal agency from stopping distribution of a model on its own. One account says it would limit state interventions to the most severe scenarios.

The catch: Critics argue the judicial gate makes blocking procedurally hard enough that agencies might not try.

In their words: An Anthropic spokesperson described the ideas as "sensible policies we believe Congress should act on."

Unconfirmed: These details come from secondary summaries of the draft. The bill's text and its sponsors have not been seen.

What's next: The notes name no sponsor or timetable. Anthropic says Congress should act.