AI Systems Transparency Act

This bill would require the Federal Trade Commission (FTC), working with the National Institute of Standards and Technology, to write rules requiring companies that build widely used generative AI models to publicly post information about them. Covered models are those used by at least 1,000,000 monthly U.S. users, or at least 100,000 if minors can create accounts. Developers would post one disclosure for consumers and one detailed disclosure for researchers, covering safety measures for minors and other vulnerable users, privacy practices, uses and limits, security, and risk testing, and would update that information over time. Similar rules would apply to companies that run these systems and to companies that significantly modify the models. Violations would be treated as unfair or deceptive practices enforced by the FTC.

Full text and official record on Congress.gov

Sponsors

• Christopher A. Coons Democrat sponsor
• 3 cosponsors