Anthropic Proposes Federal AI Bill to US Lawmakers

Anthropic, the developer of the "Claude" artificial intelligence assistant, confirmed on Friday that it is circulating a draft bill among US lawmakers designed to regulate the technology. The move comes amid a lack of federal legislation addressing risks posed by new AI systems.

Key Provisions of the Draft

According to the draft first revealed by Politico on Friday, Washington would gain the authority to block the launch or use of any AI model if there is an imminent risk of death or serious injury to more than 50 people, or damages exceeding $1 billion. Such interventions would require judicial approval.

A company spokesperson told AFP that the proposal grants the federal government the power to prevent or deter dangerous deployments of the most powerful AI models. The text also obliges leading developers to test their strongest models, maintain transparency with the public regarding test results, subject models to independent evaluations, and uphold robust security programs.

Current Regulatory Landscape

US AI policy currently relies primarily on voluntary commitments from industry players, alongside executive orders intended to support the sector's growth, which tech leaders view as an engine for the American economy. Congress, politically divided and facing intense pressure from major tech firms, has not yet passed a federal law regulating AI and remains in recess until after next month’s midterm elections.

Security concerns have intensified following recent incidents involving "AI agents," software built on AI models capable of autonomously executing sequences of tasks. Anthropic previously suspended its most advanced models for about three weeks under pressure from Washington, after the Donald Trump administration barred the company from dealing with the Department of Defense in February.

Industry Reactions and Precedents

Last month, Anthropic CEO Dario Amodei urged the sector to slow development, warning that a group of AI agents could potentially take control of the "entire internet" within six to 12 months. In contrast, Trump described repeated calls to decelerate technological progress as "negative forces."

The proposed legislation would override certain state-level laws, such as those currently in effect in California. This marks a shift from June, when Anthropic stated that in the absence of strong federal law, states should retain full authority to enact their own regulations.

Charlie Pollock of the Institute for Law and AI research center praised the proposal as "very reasonable" and superior to what other leading AI labs have offered. However, he noted in an X post that the draft appears "weaker" than the "Frontier" Act, a competing bill introduced in the House of Representatives this summer.