Anthropic has privately pitched U.S. lawmakers a draft bill to regulate advanced artificial intelligence, according to sources familiar with the matter. The proposal would establish a framework for AI safety and security, including independent evaluators registered with the Department of Commerce to assess AI models.
The bill would let authorities block a model’s release or use only if it posed a foreseeable and imminent risk of causing more than $1 billion in damages. It would also allow intervention if a model posed such a risk of killing or seriously injuring more than 50 people.
The proposal would limit which federal agencies could suspend AI development. It aims to streamline oversight and prevent regulatory overreach.