What’s the history

Originally introduced as a discussion draft in the Fall of 2023, The NO FAKES Act creates a federal intellectual property-style right protecting each person’s voice and visual likeness against unauthorized, highly realistic AI-generated “digital replicas.” The acts provision not only apply to individuals that have historically benefited from the various Right of Publicity laws promulgated in many states, but also private individuals. 

In practical terms, the bill is aimed at deepfakes, cloned voices, AI avatars, synthetic endorsements, and other commercial or public-facing uses that make it appear that a real person performed, appeared, spoke, or endorsed something when they did not.

For businesses, the core message is straightforward: if your organization uses AI-generated voices, faces, avatars, performances, testimonials, ads, social content, or entertainment assets that are identifiable as real people, you should plan for a consent-and-licensing model. The bill is not yet law, but it has bipartisan support and has advanced further than earlier versions.

Earlier versions of the bill failed to garner significant support. However, after continued negotiations that lead to some major revisions, the bill was reintroduced in May 2026.

What’s next

Currently, the NO FAKES Act of 2026 remains pending legislation and is not yet federal law. The revised version of the bill was introduced as Senate bill, S. 4591 on May 20, 2026. The revised bill was considered by the Senate Judiciary Committee and ordered reported to the full Senate on June 18, 2026. The committee-reported text was published on June 24, 2026. The next key step will be consideration by the full Senate, followed by passage of matching legislation by both chambers and signature by the President.

The bill has meaningful bipartisan support, with sponsors and co-sponsors from both parties. It also has attracted close attention from the entertainment, music, sports, advertising, technology, and online platform sectors because it would create a national framework in an area currently governed largely by a patchwork of state right-of-publicity, privacy, unfair competition, and deepfake laws.

What does the Act do

• Creates a federal voice and likeness right. The bill would give individuals a federal right to control the use of their voice and visual likeness in covered digital replicas.
• Covers living and deceased individuals. The right would apply to human beings, living or dead, and would be licensable.
• Targets highly realistic AI replicas. The definition focuses on newly created, computer-generated, highly realistic electronic representations that are readily identifiable as a real person’s voice or likeness.
• Permits civil enforcement. Unauthorized creation, distribution, publication, performance, or hosting of covered digital replicas could create civil liability, subject to statutory limitations and defenses.
• Creates platform obligations. Online services would have a notice-and-takedown framework and potential safe harbor if they respond appropriately to qualifying notices.
• Includes counter-notice protections. The 2026 version includes a counter-notification process, giving users a mechanism to contest removal of allegedly infringing digital replicas.
• Preserves important speech exceptions. The bill contains exceptions intended to protect news, documentary, commentary, criticism, parody, satire, sports broadcasts, and other First Amendment-sensitive uses.

Why it matters 

The Act is best understood as a proposed federal rule of the road for synthetic identity. If enacted, it would likely affect advertising, influencer marketing, entertainment production, music, gaming, sports media, social platforms, brand protection, customer service avatars, political advertising, and any business using generative AI to simulate a real person.

Unfortunately, even if enacted, the federal law will coexist with certain state-law rights and remedies, so state publicity and deepfake laws will remain relevant. So, businesses will still need to monitor the relevant state laws to accurately access risks.

What’s the practical view

From a trademark and brand-protection perspective, the NO FAKES Act does not replace traditional claims such as false endorsement, trademark infringement, unfair competition, right of publicity, or consumer deception. Instead, it adds a more targeted federal cause of action for AI-enabled impersonation. That is commercially significant because many harmful deepfakes do not fit neatly into existing trademark or copyright frameworks: the problem is often not misuse of a mark or copying of a protected work, but unauthorized appropriation of identity.

Businesses should treat the bill as a strong signal of where U.S. law is heading. Even before enactment, the best practice is to implement written consent, maintain records of AI-generated assets, avoid synthetic endorsements without express permission, and ensure contracts with agencies and AI vendors allocate responsibility for digital replica compliance.