U.S. CATS Act Allows AI Firms Limited Antitrust Protection for Security Coordination

Bipartisan bill grants AI firms antitrust immunity to share intel on model distillation and restrict high-risk deployments, targeting Chinese labs while barring price-fixing.

Woofun AI reports that bipartisan U.S. lawmakers introduced the "CATS Act" to provide limited antitrust protection for AI companies. The legislation permits competitors to share intelligence regarding model theft, distillation, and cyber attacks. In cases of national security risks, firms may collectively delay or restrict the deployment of high-risk models. Proponents identify China as a primary target, citing previous accusations against Chinese labs by Anthropic, OpenAI, and Google. Before imposing joint restrictions, companies must submit written notice to the U.S. Department of Justice and prove actions address security risks only. Price fixing, market allocation, monopolies, and group boycotts remain illegal.