Anthropic is said to have quietly presented its own draft AI safety and security bill to U.S. lawmakers, according to a Politico report carried by multiple outlets. Bloomberg and CNA describe the pitch as an effort to shape regulation around AI model safety.
The proposed framework would give the government authority to intervene if an AI model is expected to pose a serious, foreseeable and imminent risk. NDTV reports that the threshold for potential government action is tied to harm estimates, including more than $1 billion in damages, or the risk of killing or seriously injuring more than 50 people. Other coverage emphasizes the existence of the bill and its focus on safety oversight, without adding many additional details in the excerpts provided.
Across the reports, the main common thread is that Anthropic’s draft bill would create conditions under which authorities could block or restrict a model’s release or use. The outlets differ mainly in how much specific threshold language they include, but they broadly agree on the bill’s origin and its intended regulatory purpose.
