AI Impersonation Prevention Act of 2025

Summary

Makes it a federal crime to knowingly use artificial intelligence to impersonate a United States officer or employee without an explicit disclaimer when the impersonation produces materially false or misleading content. Violations carry fines and/or up to three years in prison; clear exceptions protect parody and satire that are disclosed as not authentic.

• Creates a federal crime for knowingly using AI to impersonate a U.S. officer or employee without an explicit disclaimer when the impersonation produces materially false or misleading content.
• Defines "artificial intelligence" to include systems or software (including generative models) that generate human-like audio, video, or text.
• Defines "impersonates" as falsely representing oneself as an identifiable person in a way likely to make viewers believe the content is authentic.
• Authorizes fines and/or up to three years imprisonment for violations.
• Protects satire and parody that clearly discloses it is not authentic as First Amendment-protected conduct.
• Contains a severability clause so remaining provisions survive if part of the law is invalidated.
• Does not provide funding, administrative rules, or an explicit effective date in the text.
• Imposes practical compliance requirements on AI creators, platforms, and distributors to use clear disclaimers when depicting federal officers or employees.