United States - Media Integrity Protection (S. 1396)

Content Origin Protection and Integrity from Edited and Deepfaked Media Act (S. 1396, 2025)

United States

RAI-US-NA-COPIFXX-2025
Proposed(Officially filed for action)
BillGovernance and OversightTransparency and DisclosureEnforcement and Penalties
Export PDF

S.1396 (2025) requires transparency for synthetic and synthetically-modified content by promoting machine-readable content provenance, supporting standards development and NIST research, mandating provenance options for commercial tools, prohibiting tampering with provenance data, and restricting nonconsensual use of provenance-tagged copyrighted works for AI training. It creates enforcement pathways including FTC jurisdiction and private civil actions.

Summary

The Content Origin Protection and Integrity from Edited and Deepfaked Media Act (S.1396), introduced April 9, 2025, aims to increase transparency, protect creators, and strengthen trust in digital media by advancing machine-readable content provenance, watermarking, detection tools, standards, and public education. The bill defines covered terms including “content provenance information,” “synthetic content,” “synthetically-modified content,” “covered content” and “covered platform.” It directs the Under Secretary of Commerce for Standards and Technology (NIST) to establish a public-private partnership to facilitate voluntary consensus-based standards and best practices for watermarking, provenance metadata, and detection metrics, and to consult with the Register of Copyrights and the Director (Under Secretary of Commerce for Intellectual Property and Director of the USPTO). NIST is instructed to run research and development programs on measurement science, robustness, and cybersecurity for provenance and detection technologies, and to launch a public education campaign within one year of enactment.

Key operational requirements include: a 2-year compliance window for commercial tools whose primary purpose is creating synthetic or substantially modifying covered content to provide users the ability to include content provenance information; reasonable security measures to make provenance machine-readable and resistant to removal; prohibitions on knowingly removing, altering, tampering with, or disabling content provenance information in furtherance of unfair or deceptive acts in commerce; an exception allowing platforms to remove or alter provenance only for narrowly tailored security research; and a prohibition on the commercial use of copyright-protected covered content that has attached provenance information (or from which provenance has been removed in violation of the Act) for the purpose of training AI systems without the express, informed consent of the copyright owner and compliance with any terms, including compensation terms. The bill treats violations as unfair or deceptive acts under the Federal Trade Commission Act, granting enforcement authority to the FTC (with powers, penalties, and procedures mirroring the FTC Act), provides for state attorney general parens patriae actions, and preserves a private civil cause of action for copyright owners whose provenance information is tampered with or whose works are used in violation.

Relief available includes injunctive and declaratory relief, compensatory damages, restitution and reasonable litigation costs including attorney fees. The statute of limitations for private suits under the bill is 4 years from discovery. The bill also instructs NIST to coordinate prize challenges with DARPA and NSF to spur advances in detection and anti-tamper measures and directs consultation among federal intellectual property and standards offices. The Act contains a construction clause preserving other copyright owner rights and a severability clause. Overall, S.1396 blends standards-setting and technical support with regulatory prohibitions and enforcement tools to address the provenance, integrity, and lawful use of synthetic and edited media.

Full article

Read full text ↗

Overview

The Content Origin Protection and Integrity from Edited and Deepfaked Media Act (S.1396, 119th Congress) is a bipartisan bill introduced April 9, 2025 by Sen. Maria Cantwell (with cosponsors Sen. Marsha Blackburn and Sen. Martin Heinrich) to increase transparency and protect creators against unauthorized use of copyrighted works in the era of algorithmically generated and modified content. The bill directs the Under Secretary of Commerce for Standards and Technology to establish a public-private partnership and to develop voluntary, consensus-based standards and guidelines for machine-readable content provenance information (provenance metadata), watermarking, and detection of synthetic and synthetically-modified content. It requires commercial tool providers to offer the ability to embed provenance metadata and imposes civil and administrative enforcement mechanisms, treating violations as unfair or deceptive acts under the Federal Trade Commission Act and enabling private causes of action for affected rights-holders. The overall intent is to foster measurement science, interoperability, and public education while deterring tampering and nonconsensual commercial exploitation of provenance-tagged works.

Definitions

S.1396 establishes specific definitions crucial to scope and enforcement: "content provenance information" is defined as state-of-the-art, machine-readable data documenting a content item’s origin and history; "covered content" references works of authorship under 17 U.S.C. §102; "covered platform" is a website/application available in the U.S. meeting thresholds ($50,000,000 annual revenue or 25,000,000 monthly active users for at least 3 of the prior 12 months); "synthetic content" and "synthetically-modified content" capture algorithm-generated and significantly algorithm-altered materials; "deepfake" is content that appears authentic and creates a false impression; "Under Secretary" refers to the Under Secretary of Commerce for Standards and Technology; "Director" refers to the Under Secretary of Commerce for Intellectual Property and Director of the USPTO. These definitions anchor obligations, exceptions, and enforcement tools.

Governance and Institutional Framework

The bill assigns a central coordinating role to the Under Secretary of Commerce for Standards and Technology (NIST) to establish a public-private partnership for development of voluntary consensus-based standards and best practices, and to lead NIST research programs on measurement, testing, and cybersecurity for provenance and detection tools. The Under Secretary must consult with the Register of Copyrights and the Director of the USPTO in standards development and public education activities. The Act explicitly references coordination with the Defense Advanced Research Projects Agency (DARPA) and the National Science Foundation (NSF) for grand challenges and prizes. For enforcement, the Federal Trade Commission (FTC) is given primary administrative enforcement authority under the FTC Act analog (section 7), and State attorneys general retain parens patriae authority; private civil suits are preserved for copyright owners and other affected parties. The bill thus spans standards-setting, federal research, agency enforcement, and private rights of action to create a multi-institutional governance approach. Key institutional players include NIST (NIST), the U.S. Copyright Office (U.S. Copyright Office), the U.S. Patent and Trademark Office (USPTO), the Federal Trade Commission (FTC), and research agencies such as DARPA (DARPA) and NSF (NSF).

Key Focus Areas

S.1396 focuses on four interlocking domains. First, standards and interoperability: convening public-private partnerships to build machine-readable provenance standards, watermarking guidance, detection metrics, and evaluation tools, including guidance on data-use transparency for model training. Second, technical robustness and anti-tamper measures: directing NIST research on measurement science, test methods, and cybersecurity protections to make provenance metadata resistant to removal or alteration. Third, industry obligations: requiring commercial providers of tools primarily used to create synthetic or substantially modify covered content to offer provenance embedding options and to implement reasonable security measures to keep provenance machine-readable and bound to content. Fourth, legal prohibitions and consent: making it unlawful to knowingly remove or tamper with provenance in furtherance of unfair or deceptive commerce and prohibiting the commercial use of provenance-tagged covered content for AI training absent express, informed consent of the copyright owner and compliance with usage/compensation terms. The bill also contemplates exception carve-outs for narrowly tailored security research and provides enforcement mechanisms via the FTC, state attorneys general, and private civil suits to deter and remedy abuses.

Implementation Framework

Operational steps in the bill include formation of the public-private partnership at NIST, issuance of voluntary standards and best practices, NIST-conducted research programs, and a public education campaign to be launched within 1 year of enactment. Tool providers and covered-platform actors have a 2-year period after enactment to implement user options for embedding content provenance information and to put in place security measures to resist tampering. NIST is directed to coordinate grand challenges and prizes with DARPA and NSF to accelerate detection and anti-tamper solutions, while standards development is to take into account input from intellectual property offices (USPTO and the U.S. Copyright Office). Covered platforms (per the thresholds) are specifically restricted from separating provenance from content so that users cannot access the provenance, except where removal is strictly necessary and proportionate for security research. The bill relies primarily on voluntary consensus-based standards but converts several baseline expectations into enforceable prohibitions tied to FTC authority and civil liability.

Monitoring and Evaluation

Monitoring is multi-layered: NIST-led research will generate metrics and test methods for detection and provenance robustness; the public-private partnership will facilitate evaluation criteria and red/blue-teaming methodologies to assess tool efficacy; the FTC will monitor marketplace practices and enforce prohibitions as unfair or deceptive acts. State attorneys general and private plaintiffs create additional monitoring/discovery incentives by enabling litigation and recovery for rights-holders. The bill’s requirement for metrics, guidelines, and grand challenges aims to create living evaluation processes to measure detection performance, anti-tamper resilience, and adoption of provenance practices across industry sectors. Periodic public outputs (guidelines, test results, prize outcomes) are expected to inform iterative updates to standards.

Penalties, Liability, and Appeals

Violations are treated as unfair or deceptive acts under the FTC Act, subjecting violators to the FTC’s remedies and penalties. The bill authorizes the FTC to enforce and seek remedies under the same powers as for other FTC Act violations. States’ attorneys general may bring parens patriae actions to enjoin violations, compel compliance, and obtain damages, restitution, or other relief. Copyright owners and related parties can bring private civil actions against persons or platforms that remove, alter, tamper with, or disable content provenance information or that use covered content for training in violation of the consent requirement. Courts may award declaratory or injunctive relief, compensatory damages, and reasonable litigation expenses (including attorney’s fees). The statute of limitations for private suits under this Act is four years from discovery. The Act preserves other legal rights and provides severability to safeguard enforceable provisions if parts are invalidated.

Relationship to Other Instruments

S.1396 is designed to work alongside existing federal statutes and agencies. It explicitly references the Federal Trade Commission Act for enforcement; cross-references copyright law (title 17 U.S.C.) by protecting rights of copyright owners and requiring consent/compensation where required by copyright holders; and builds on the National Artificial Intelligence Initiative Act definition of AI. The bill complements ongoing regulatory and legislative efforts on nonconsensual deepfakes (e.g., NCII-focused measures), platform safety rules, and privacy law initiatives. It seeks to avoid preemption or impairment of copyright owners’ existing statutory rights while adding provenance and anti-tamper obligations and new civil remedies targeted specifically at provenance tampering and nonconsensual commercial training uses.

International Alignment

The bill promotes standards-based approaches and voluntary consensus models that can interoperate with international provenance and watermarking efforts. By directing NIST to develop open standards and measurement science, S.1396 aims to facilitate cross-border technical alignment with partners and standards bodies (ISO, W3C, industry consortia) and to support U.S. participation in international harmonization. The consent and compensation provisions are primarily domestic in reach but may influence global platform policies and data-sharing agreements. The bill’s research and prize approaches (with DARPA and NSF) are intended to advance detection tools that can be shared internationally with allies and standards communities to counter malicious deepfakes and preserve cross-border content integrity.

Implementation Timeline

MilestoneDate/Deadline
Bill introduced in Senate2025-04-09
Public education campaign (NIST)Within 1 year of enactment
Standards development & public-private partnership establishedImmediate post-enactment; phased outputs (ongoing)
Commercial tools and covered content tool compliance2 years after enactment
Private civil action statute of limitations4 years from discovery

Sources and References

SourceType
S.1396 — Congress.gov (bill text and page)Primary Source
S.1396 — PDF (Congress.gov)Primary Source

Requirements for a company

What an organisation has to do under United States - Media Integrity Protection (S. 1396), at a glance. Not legal advice — the table below gives the provision and deadline for each item.

Not yet in force (Proposed). These requirements apply once the instrument takes effect and may change before then.

Must do

5
  • Offer users the ability to embed machine-readable content provenance information.Commercial providers of tools creating or modifying covered content.
  • Implement reasonable security measures to keep provenance machine-readable and bound to content.Commercial providers of tools creating or modifying covered content.
  • Do not knowingly remove or tamper with content provenance information.Any person or platform.
  • Obtain express, informed consent before commercially using provenance-tagged content for AI training.Any person or platform commercially using provenance-tagged covered content for AI training.
  • Do not separate provenance from content, preventing user access, unless for security research.Covered platforms (meeting revenue or user thresholds).

Must not do

0

Nothing in this category.

Should do

1
  • Engage with the NIST public-private partnership for standards development.Industry stakeholders involved in synthetic content.

Should not do

0

Nothing in this category.

Who must do what

The obligations under United States - Media Integrity Protection (S. 1396), most serious first. Not legal advice — verify against the official text before relying on it.

#WhoRequirementBy whenWhereSeverity
1Commercial providers of tools creating or modifying covered content.Offer users the ability to embed machine-readable content provenance information.
requiring commercial providers of tools primarily used to create synthetic or substantially modify covered content to offer provenance embedding options
2 years after enactmentCritical
2Commercial providers of tools creating or modifying covered content.Implement reasonable security measures to keep provenance machine-readable and bound to content.
implement reasonable security measures to keep provenance machine-readable and bound to content.
2 years after enactmentCritical
3Any person or platform.Do not knowingly remove or tamper with content provenance information.
making it unlawful to knowingly remove or tamper with provenance in furtherance of unfair or deceptive commerce
Critical
4Any person or platform commercially using provenance-tagged covered content for AI training.Obtain express, informed consent before commercially using provenance-tagged content for AI training.
prohibiting the commercial use of provenance-tagged covered content for AI training absent express, informed consent
Critical
5Covered platforms (meeting revenue or user thresholds).Do not separate provenance from content, preventing user access, unless for security research.
Covered platforms... are specifically restricted from separating provenance from content so that users cannot access the provenance
Implementation FrameworkCritical
6Industry stakeholders involved in synthetic content.Engage with the NIST public-private partnership for standards development.
The bill directs the Under Secretary of Commerce for Standards and Technology to establish a public-private partnership and to develop voluntary, consensus-based standards
ongoingRecommended

© Regulations.AI · updated on 13-Jun-2026