Anthropic secretly lobbied Congress: government could only halt a model if it faces over $1 billion in losses or risks 50 casualties

Beating AI Brief: Anthropic is privately lobbying U.S. lawmakers on a draft frontier AI safety bill; according to Politico, the draft aims to set strict thresholds for government intervention in AI models. The draft stipulates that the government may seek to halt a model only when it faces specific, foreseeable and imminent risks—potentially causing more than $1 billion in losses, or leading to 50 or more deaths or serious injuries. The regulatory triggers are limited by the draft to three categories of risk: AI models being used for the development or proliferation of bioweapons, models escaping developer control, and models enabling cyberattacks. The bill also requires AI developers to conduct transparency testing and undergo independent third-party evaluations; if an evaluating body identifies relevant risks, it must report to the U.S. Department of Commerce within 72 hours, and the Commerce Department may then petition a federal court to suspend the model's operations, with the court making the final ruling. Additionally, the draft would prohibit other U.S. government agencies and officials from independently halting the development, deployment, or internal use of AI models. Anthropic said the proposal stems from its advanced AI framework released earlier this year, which aims to require developers to undergo safety evaluations while giving the government the ability to block the deployment of dangerous models. The draft is currently still in the lobbying stage and has not become law. If the relevant provisions are ultimately passed, the U.S. government's authority to take emergency intervention against frontier AI models could be significantly constrained.