NO FAKES is a backronym for the Nurture Originals, Foster Art, and Keep Entertainment Safe Act. As described in a one-pager published on Senator Chris Coons' website, Coons co-sponsors the bill alongside Senators Marsha Blackburn, Amy Klobuchar, and Thom Tillis, it's a bipartisan bill meant to protect the voice and visual likeness of individuals from unauthorized AI-generated recreations.

The bill started as a discussion draft in 2023, was formally introduced in 2024, reintroduced in 2025, and reintroduced again in May 2026 as a stronger, more refined version (S. 4591 and its House companion, H.R. 8915). That 2026 version is currently the one moving through Congress.

Read also: Senate committee advances NO FAKES Act

 

Why does it exist?

Currently, there's no federal right of publicity in the United States. Protection against someone misusing your name, image, or voice depends on a patchwork of state laws, which vary and often weren't written with AI in mind. The U.S. Copyright Office flagged this gap directly, recommending in a 2024 report that Congress step in and create a federal law addressing unauthorized "digital replicas," since existing protections were too narrow and inconsistent to keep up with the technology.

The Coons office one-pager points to three real-world cases that show the problem:

  • "Heart on My Sleeve" - an AI-generated track built with AI replicas of Drake and The Weeknd's voices, which racked up hundreds of thousands of listens on YouTube, Spotify, and other platforms before it was identified as fake and taken down.
  • The Tom Hanks dental ad - an AI-generated version of Tom Hanks appeared in advertisements for a dental plan he never actually appeared in or endorsed.
  • The Baltimore principal incident - a Maryland high school principal was framed as having made racist remarks via an AI-generated recording of his voice, remarks he never actually said.

 

Where ‘Right of Publicity’ comes from

According to a legal analysis published by the Columbia Undergraduate Law Review (CULR), the right of publicity isn't new. CULR traces its roots to two landmark cases, that is, Lugosi v. Universal Pictures (1979) and Midler v. Ford Motor Co. (1988), which together established that intentionally imitating someone's likeness for commercial gain is unlawful. Notably, CULR points out that those early rulings held this right was not inheritable, meaning it didn't pass on to a person's heirs after death. It was only later that individual states began passing their own laws making the right of publicity a kind of transferable property that could be inherited by an estate, treating a celebrity's persona as something with real pecuniary value.

CULR's analysis highlights two ongoing legal disputes that show why courts and lawmakers alike see this as an urgent issue:

  • Main Sequence v. Dudesy (2024) - The estate of comedian George Carlin sued the podcast Dudesy over an AI-generated special that used a deepfake replica of Carlin's voice. The case settled, with the defendant agreeing to pull the special and refrain from using Carlin's likeness again without permission. CULR notes, however, that the settlement didn't establish whether that outcome would bind other third parties, leaving the broader legal question unresolved.
  • Lehrman v. Lovo - An ongoing case in which voice actors Paul Lehrman and Linnea Sage allege that AI startup Lovo used their voices without authorization in a generative text-to-speech product. As CULR frames it, the outcome of this case will be important for understanding how unauthorized digital replicas get treated under existing law while Congress decides whether to act.

 

What the Bill would actually do

The NO FAKES Act would create a new federal intellectual property right. Every individual, famous or not, would own a licensable right in their own voice and visual likeness. The one-pager from Senator Coons' office summarizes the legislation's core provisions this way as,

“The NO FAKES Act would address the use of non-consensual digital replications in these kinds of audiovisual works, images, or sound recordings. The legislation would:

  • Hold individuals or companies liable if they produce an unauthorized digital replica of an individual in a performance;
  • Hold platforms liable for hosting an unauthorized digital replica if the platform has actual knowledge of the fact that the replica was not authorized by the individual depicted;
  • Exclude certain digital replicas from coverage based on recognized First Amendment protections; and
  • Largely preempt State laws addressing digital replicas to create a workable national standard.”

The 2026 version now moving through Congress has expanded on that original framework:

  • A DMCA-style notice-and-takedown system. Rights holders can send a takedown notice to get unauthorized content removed. Importantly, the 2026 version added something earlier drafts lacked which is a formal counter-notification process. If you post something and believe it was wrongly taken down you can file a counter-notice. If the original complainant doesn't sue within 14 days, the content goes back up. Anyone who knowingly files a bogus counter-notice faces steep penalties, up to $25,000 or actual damages.
  • Broader First Amendment carve-outs. Beyond the "recognized First Amendment protections" mentioned in the original one-pager, the bill now specifically excludes uses like news reporting, commentary, criticism, and parody. It also now exempts nonprofit libraries, archives, and accredited nonprofit educational institutions acting without commercial purpose.
  • Damages. Violators could face statutory damages, with recent versions setting a range between $5,000 and $25,000 per violation, or actual damages plus any profits earned from the unauthorized replica. Platforms that willfully ignore removal requests could face penalties as high as $750,000.
  • Life plus 70. The publicity right wouldn't disappear when someone dies. It's inheritable, and under the current draft, it can last up to 70 years after an individual's death, provided the right is actively used or renewed.

Senator Coons' office frames the bill's overall approach as one that "balances the need to protect individuals and creators, First Amendment considerations, and fostering U.S. leadership and innovation in AI."

 

The First Amendment debate

Not everyone agrees the bill's free speech carve-outs go far enough. CULR's analysis raises a pointed concern that earlier drafts of the bill defined "digital replica" broadly enough that it could sweep up more than just deepfakes explicitly created to deceive, potentially touching lawful speech like citizen journalism or commentary, depending on how the language is interpreted. CULR draws a comparison to two Supreme Court rulings where sweeping content restrictions aimed at protecting people online were struck down as unconstitutional. Reno v. ACLU (1997), which rejected provisions of the Communications Decency Act as an overly broad restriction on internet speech, and Ashcroft v. ACLU(2004), which struck down similar restrictions under the Child Online Protection Act. The takeaway in that analysis is that legislation this broad has a track record of running into First Amendment trouble if it isn't drafted narrowly enough.

Later drafts of NO FAKES, including the 2026 version now moving through Congress, added more explicit exemptions for news reporting, commentary, criticism, and parody, which addresses at least some of this concern.

 

Who's on board and who isn't

The coalition backing NO FAKES includes SAG-AFTRA, the RIAA, Universal Music Group, the Recording Academy, the Motion Picture Association universe of allies, and tech companies including OpenAI, Google, Amazon, Adobe, and IBM. Supporters argue it strikes the right balance between protecting people from deepfakes and preserving room for AI innovation and free expression.

However, digital rights groups like the Electronic Frontier Foundation (EFF) and Public Knowledge have raised concerns about how the property right could be transferred or licensed away. One critique, because the right is licensable, it could theoretically be signed away in the fine print of an app's terms of service, or seized in a bankruptcy or debt collection proceeding meaning someone could lose control over their own likeness rights without fully realizing it.

EFF's objections, laid out in a June 2025 analysis by Katharine Trendacosta and Corynne McSherry, argue that the newer version of the bill doesn't just create a licensing right, it builds an aggressive enforcement machine around it. Their analysis points out that the bill extends liability beyond unauthorized replicas themselves to the tools used to produce them, and while that liability is limited to tools "primarily designed" for making unauthorized replicas, EFF warns developers could still face serious exposure from a bare allegation alone, well before any wrongdoing is proven. They also argue the DMCA-style takedown system goes further than existing copyright law by effectively mandating that platforms filter out flagged content to prevent re-upload, which they warn could produce the same over-flagging problems already seen with copyright filters. They also state that someone could obtain a subpoena from a court clerk to unmask an anonymous user's identity based on nothing more than an allegation, a power EFF says has already been used to pressure or silence critics in copyright disputes. They further contend the added compliance burden would fall hardest on smaller platforms trying to compete with entrenched tech giants, some of which are already backing the bill, and argue Congress should slow down and evaluate the recently passed Take It Down Act before layering on new rules. Summing up their view, Trendacosta and McSherry state, "This bill is a solution in search of a problem."

 

FAQs

Is the NO FAKES Act currently the law?

Not yet, it's still a bill working its way through Congress.

 

Who would be responsible for enforcing this right, the government or individuals?

It would function as a private property right, meaning individuals (or their estates) would enforce it through civil lawsuits, not a government agency.

 

What's the usual relationship between federal and state right-of-publicity laws?

Federal law can preempt (override) conflicting state laws, but the exact scope of that preemption is often left to courts to interpret.

 

Are AI voice assistants like Siri or Alexa affected by publicity-rights laws?

Not really, since those use licensed, purpose-built synthetic voices rather than unauthorized copies of a specific real person.