Anthropic pitches bill to limit federal blocks on AI models

Anthropic has presented US lawmakers with a draft bill that would let federal authorities seek to block dangerous AI models only in the most severe cases, according to TechSpot. The proposal is not law.

The company said the draft reflects "sensible policies we believe Congress should act on," according to an Anthropic spokesperson. TechSpot attributed the original reporting to Politico.

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

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

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.

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.

Critics argue the court requirement makes blocking a model procedurally hard enough that agencies might not attempt it.

The details come from secondary summaries of the draft. The bill's text and its sponsors have not been seen, and it is not yet known whether any lawmaker will introduce it.