S. 1396 — Content Origin Protection and Integrity from Edited and Deepfaked Media Act of 2025

This section states the sense of Congress that artificial intelligence (AI) systems lack visibility into operations, transparency on training data, and consensus-based standards for development and deployment; that assessing the nature, origins, and authenticity of algorithmically generated or modified digital content is increasingly difficult; that these deficiencies harm the public and content creators (i.e., journalists, publishers, broadcasters, and artists) whose works train AI and compete with synthetic content; and that consensus-based standards would mitigate these effects, spur innovation, and position the United States to lead in AI.

This section defines 12 terms for purposes of this title, including "artificial intelligence" (as defined in 15 U.S.C. 9401); "artificial intelligence blue-teaming" (i.e., operational vulnerability evaluations and mitigation techniques for AI systems); "artificial intelligence red-teaming" (i.e., structured adversarial testing to identify AI system risks and vulnerabilities); "content provenance information" (i.e., machine-readable information on the origin and history of digital content); "covered content" (i.e., digital representations of works of authorship under 17 U.S.C. 102); "covered platform" (i.e., a website, internet application, or mobile application available to U.S. users that generates at least $50 million in annual revenue or has at least 25 million monthly active users for at least 3 of the prior 12 months); "deepfake"; "synthetic content" and "synthetically-modified content"; "watermarking"; "Director" (i.e., Under Secretary of Commerce for Intellectual Property and Director of the USPTO); and "Under Secretary" (i.e., Under Secretary of Commerce for Standards and Technology).

This section directs the Under Secretary to establish a public-private partnership to facilitate development of standards for content provenance information technologies (i.e., tools tracking media origin and alterations) and detection of synthetic content (i.e., AI-generated) and synthetically-modified content. The partnership must address (1) guidelines and voluntary consensus-based standards and best practices for watermarking, provenance information, detection tools (for images, audio, video, text, and multimodal content), data use in AI training, and synthetic media transparency; (2) guidelines, metrics, and practices to evaluate detection and labeling tools, including AI red-teaming (adversarial testing) and blue-teaming (defensive testing); and (3) grand challenges and prizes, coordinated with the Defense Advanced Research Projects Agency and National Science Foundation, for detection, labeling, and cybersecurity countermeasures against tampering with tools, watermarks, or provenance information. In developing these standards, the Under Secretary must consult with the Register of Copyrights and the Director.

This section directs the Under Secretary of Commerce for Standards and Technology to (1) carry out a research program to advance measurement science, standards, and testing for technologies detecting, watermarking, and tracking provenance of synthetic content and synthetically-modified content (i.e., AI-generated or altered media such as deepfakes), as well as cybersecurity protections and countermeasures against tampering with those technologies; and (2) not later than one year after enactment, in consultation with the Register of Copyrights and the Director, conduct a public education campaign on such content, watermarking, and provenance information.

This section requires that, beginning two years after enactment, any person who commercially distributes a tool primarily designed to create synthetic content or synthetically-modified content (i.e., AI-generated or AI-altered digital media) to (1) enable users to attach content provenance information (CPI) indicating such status, per standards under section 4; and (2) if users opt in, implement feasible security measures to make the CPI machine-readable and resistant to removal, alteration, or separation. It further requires the same obligations for tools primarily designed to create or substantially modify covered content (i.e., digital depictions of identifiable individuals). This section prohibits (1) any person from knowingly removing, altering, tampering with, or disabling CPI to further unfair or deceptive commercial practices; (2) covered platforms from doing so or separating CPI from content, except as necessary, proportionate, and limited to security research; and (3) commercial use of covered content with attached CPI—or from which CPI was unlawfully removed under these rules—to train AI systems or generate synthetic/synthetically-modified content without the copyright owner's express, informed consent and compliance with applicable terms, including compensation.

This section establishes enforcement for violations of the Act. Subsection (a) treats such violations (including of regulations promulgated under the Act) as violations of a rule defining unfair or deceptive acts or practices under section 18(a)(1)(B) of the FTC Act (15 U.S.C. 57a(a)(1)(B)), authorizes the FTC to enforce them using all applicable FTC Act authorities, penalties, privileges, and immunities (15 U.S.C. 41 et seq.), and preserves other FTC authorities. Subsection (b) authorizes state attorneys general (or other authorized state officials) to bring civil parens patriae actions in federal court to enjoin violations, compel compliance, obtain damages or restitution for residents, or seek other relief; requires notice to the FTC (with an exception for infeasible cases) and allows FTC intervention and appeal rights; preserves state investigatory powers; prohibits state actions against defendants named in a pending FTC civil or administrative action; and specifies venue, service of process, and a damages offset for any concurrent private awards under subsection (c). Subsection (c) authorizes owners of covered content with attached or associated content provenance information to bring civil actions against persons or covered platforms for removing, altering, tampering with, or disabling such information (in violation of section 6(b)(1) or (b)(2)) or for misuse of such content (in violation of section 6(c)), with prevailing plaintiffs eligible for declaratory or injunctive relief, compensatory damages, and reasonable litigation costs including attorney fees; provides a four-year statute of limitations from discovery of the violation; and applies to government entities.

This section provides a severability clause, stating that if any provision of the Act or its amendments is held invalid or unenforceable, the remaining provisions and amendments remain in effect.