Anthropic proposes law regulating AI to US lawmakers

Anthropic, the creator of AI assistant Claude, has introduced a draft bill to US lawmakers, aiming to regulate artificial intelligence (AI) and address the lack of federal legislation on the technology's risks. According to the draft, the bill could block or halt a model's release or use if there is an imminent risk of killing or severely injuring more than 50 individuals or causing damages exceeding $1 billion, with a judge's approval.

The spokesperson for Anthropic stated that the proposal "grants the federal government the ability to block or deter dangerous deployments of the most powerful models." It also requires frontier developers to test advanced models, be transparent with the public, submit them to independent evaluation, and maintain a robust security program.

AI policy in the US has largely relied on voluntary industry pledges and executive orders under President Donald Trump, with Congress, heavily influenced by Big Tech, yet to pass a federal AI law. Security concerns have risen recently due to AI agents that have escaped developers' control, causing incidents. Anthropic CEO Dario Amodei has called for a slowdown in development, fearing AI agents could take control of the entire internet within six to twelve months.

The proposed legislation would also supersede certain state laws, such as those in California. Anthropic views the proposal as "pretty reasonable" but considers it "weaker" than the FRONTIER Act, a competing bill introduced in the House of Representatives.