What is the current status of the NO FAKES Act in 2026 and how does it affect AI trademark and likeness rights?

NO FAKES Act 2026: Current Status and Practical Implications for AI Trademark and Likeness Protection

The NO FAKES Act (National Obligations to Address Deepfakes and Knowledgeable Exploitation of Synthetics) has moved through the U.S. legislative process in 2026 with notable momentum but remains unsigned into law as of September 1, 2026. Originally introduced by Rep. Salazar (D-TX) alongside Senators Blackburn (R-TN) and Coons (D-DE), the bill cleared the Senate Judiciary Committee in July 2026 with unanimous bipartisan support, marking a rare moment of consensus on AI governance. However, it has not yet reached a full floor vote in either chamber. The House version, reintroduced in March 2026, includes expanded protections for libraries, researchers, and satire, responding to criticism from groups like the Electronic Frontier Foundation (EFF) and Cato Institute, who warned the original draft could chill free speech. As of late August 2026, the bill is in a conference committee phase, with lawmakers negotiating final language between the House and Senate versions.

Also worth reading: How does trademark registration for voice likeness work and protect against AI deepfakes in 2026? · What is the current pricing structure for AI trademark review services and how much does it cost to use USPTO AI tools in 2026? · What are the current trademark enforcement automation trends in 2026 and how do they impact brand protection strategies?

For trademark and AI practitioners, the NO FAKES Act introduces a federal right of publicity for voice, likeness, and identity — a significant shift from the current patchwork of state laws. This would allow individuals to sue for unauthorized use of their digital likeness in AI-generated content, including deepfakes, synthetic media, and AI training datasets. The bill also grants enforcement rights to estate representatives posthumously, a provision that has drawn both support from celebrity estates and concern from AI researchers who argue it could impede model training. The International Trade Commission (ITIF) raised specific concerns in June 2026 about the bill’s impact on video games and interactive media, suggesting exemptions for non-commercial and transformative uses.

How the NO FAKES Act Works: Legal Mechanisms and Enforcement Pathways

The NO FAKES Act establishes a civil cause of action for unauthorized commercial use of a person’s “digital replica” — defined broadly to include voice, face, body, and even distinctive mannerisms captured via AI. The law would be administered by the Federal Trade Commission (FTC) and private right of action, with statutory damages ranging from $1,000 to $100,000 per violation, plus injunctive relief. Critically, the bill includes a “safe harbor” for use in news, satire, documentary, or educational content, provided the use is not misleading and does not imply endorsement. This was added in response to EFF’s warning that the original text could criminalize political satire or parody.

Enforcement would occur through both federal courts and administrative complaints filed with the FTC. The FTC would maintain a registry of individuals who opt in to protection, similar to trademark registration, though the bill does not require registration for enforcement. The law would preempt conflicting state laws, creating a uniform national standard — a major shift from the current 35+ state right of publicity statutes, many of which are narrower in scope and shorter in duration. For AI companies, this means compliance will require robust consent management systems, especially for training data scraped from public sources. The bill also mandates that platforms hosting user-generated AI content implement “notice and takedown” mechanisms akin to DMCA, with liability shields for good-faith compliance.

Practical Steps for Trademark and AI Practitioners in 2026

For trademark attorneys and AI compliance officers, the NO FAKES Act’s trajectory demands proactive preparation. First, audit existing AI training datasets for any use of identifiable individuals without consent. This includes not just images but also voice clips, video footage, and even text prompts that could generate synthetic likenesses. Second, implement consent management platforms that track permissions for digital replica use, especially for clients in entertainment, sports, or influencer marketing. Third, review contracts with AI vendors to ensure clauses address NO FAKES liability, including indemnification for third-party claims.

For brands considering AI-generated spokespeople or virtual influencers, the bill introduces a new layer of clearance: written consent from the individual whose likeness is being replicated, or proof that the AI model was trained on public domain or licensed data. This is particularly relevant for voice cloning — the Backstreet Boys’ recent trademark filing for their voices (noted by JD Supra in 2026) signals a trend toward treating vocal signatures as intellectual property. Practitioners should also monitor the FTC’s forthcoming rulemaking on “digital replica disclosures,” which may require labeling of synthetic media in advertising. Finally, engage with local representatives: the bill’s conference committee is accepting public comments, and industry input could shape exemptions for non-commercial AI research.

Comparison: NO FAKES Act vs. Alternative Regulatory Frameworks

The NO FAKES Act is more comprehensive than existing state laws but less prescriptive than the EU AI Act, which categorizes AI systems by risk tier. Unlike the EU approach, NO FAKES does not mandate algorithmic transparency or bias audits, focusing instead on individual rights. Voluntary standards like C2PA (Coalition for Content Provenance and Authenticity) offer technical solutions for media authentication but lack legal teeth. For global businesses, the NO FAKES Act would require harmonizing compliance across jurisdictions, particularly where EU and U.S. laws diverge on data privacy and personality rights.

Common Mistakes and Misconceptions About NO FAKES

One widespread misconception is that the NO FAKES Act criminalizes all AI-generated content. In reality, it targets unauthorized commercial use, not creation. Educational tools, research models, and non-commercial satire are exempt, provided they do not mislead consumers. Another error is assuming the bill applies only to “deepfakes” — it covers any synthetic replica, including benign uses like AI-generated audiobooks or virtual assistants. Businesses often overlook the posthumous protection clause: estates can enforce rights for decades after death, complicating legacy AI training data.

A critical mistake is failing to distinguish between trademark and right of publicity. While trademarks protect brand identifiers (logos, slogans), NO FAKES protects personal identity. A company can trademark a fictional character but still need consent to use a real person’s likeness in AI ads. Additionally, many assume the bill’s safe harbor for satire is absolute — it is not. The EFF’s analysis warns that context matters: a satirical use implying endorsement (e.g., fake political ads) could still be actionable. Finally, some practitioners believe the bill is stalled; while true as of September 2026, the unanimous Senate committee vote suggests strong likelihood of passage before year-end.

When to Act: Timeline and Cost Considerations

Organizations should begin compliance planning now, anticipating final passage by Q4 2026. The bill’s effective date is likely 180 days after enactment, giving businesses roughly six months to adapt. Costs vary: small businesses might spend $5,000–$15,000 on consent audits and legal review, while large enterprises could face $100,000+ in compliance infrastructure. AI startups should budget for “right of publicity clearance” as a line item in due diligence, especially if training on public data. The FTC’s forthcoming rulemaking may impose additional disclosure requirements, adding to costs.

For trademark attorneys, the key action is to advise clients on “AI readiness”: reviewing existing licenses for training data, drafting NO FAKES-compliant consent forms, and educating marketing teams on synthetic media labeling. The bill’s passage will also trigger a surge in right of publicity registrations, similar to trademark filings — early adopters may gain competitive advantage by securing rights before the flood. Finally, monitor the conference committee’s output; amendments could narrow or expand exemptions, affecting strategy.

Sources and Further Reading

• House.gov: Rep. Salazar’s NO FAKES Act Advances Out of Senate Judiciary Committee (July 2026)
• Electronic Frontier Foundation: The NO FAKES Act Could Silence Satire, Commentary, and News (2026)
• Cato Institute: NO FAKES, Real Censorship: How Congress Could Muzzle AI Speech (2026)
• ITIF: The NO FAKES Act Needs Changes to Protect Video Games (June 11, 2026)
• IPWatchdog: NO FAKES Reintroduced with More Protections for Libraries and Researchers (2026)
• Variety: More Than 16,000 Sign SAG-AFTRA Letter Over NO FAKES Act (2026)
• Billboard: AI Deepfake Law NO FAKES Act Advances in Congress (2026)
• JD Supra: The Backstreet Boys Are Trademarking Their Voices (2026)
• World Trademark Review: Exclusive ‘.ai’ Risk Research; NO FAKES Revived (2026)

FAQ

Q: Has the NO FAKES Act been signed into law as of September 2026? A: No. As of September 1, 2026, the bill has passed the Senate Judiciary Committee unanimously and is in conference committee between House and Senate versions. It has not yet reached a floor vote or received presidential signature.

Q: Does the NO FAKES Act apply to non-commercial AI projects like open-source models? A: The current draft includes exemptions for research, education, and non-commercial use, but the boundaries are still being negotiated. The ITIF has advocated for clearer protections for open-source developers, and the conference committee may refine these clauses.

Q: How does the NO FAKES Act interact with existing state right of publicity laws? A: The bill would preempt state laws, creating a uniform federal standard. This means individuals would no longer need to file in multiple states; enforcement would occur through federal courts or the FTC, with damages set by the federal statute.

Q: What is the cost of complying with the NO FAKES Act for small businesses? A: Estimated costs range from $5,000 to $15,000 for initial audits and consent management systems, depending on the volume of AI training data and existing legal infrastructure. Ongoing compliance costs may include monitoring and renewal fees.

Q: Can estates enforce NO FAKES rights after a person’s death? A: Yes. The bill includes posthumous protection, allowing estates to sue for unauthorized use of a deceased individual’s digital replica. The duration is extendable, though specific terms are under debate in the conference committee.

Quick Facts

• Category: Federal AI legislation (right of publicity)
• Timeline: Senate committee approval July 2026; conference committee ongoing; potential passage Q4 2026
• Cost: $5,000–$100,000+ for compliance, varying by organization size
• Best for: Trademark attorneys, AI companies, entertainment brands, and individuals seeking identity protection

Follow-up Keyword

NO FAKES Act compliance checklist 2026