The NO FAKES Act, a proposed federal law that would ban AI-powered voice cloning and deepfakes, passed a key Congressional hurdle Thursday (June 18) as it was approved by a Senate committee.

The bill — which now has support from major music companies, movie studios, tech giants and entertainment unions — was endorsed by the Senate Judiciary Committee by unanimous vote and advances into the full Senate.

If enacted, NO FAKES (Nurture Originals, Foster Art, and Keep Entertainment Safe) would ban digital replicas of someone’s voice or visual likeness, a sweeping new federal prohibition aimed at combating the massive rise in fake videos and voice cloning.

AI has made it far easier to mimic voices and likenesses, flooding the internet with misleading content and leaving victims with little legal recourse. A person’s likeness has historically been guarded via so-called publicity rights, but only under a patchwork of state laws that are mostly designed to prevent unauthorized commercial advertisements. Existing federal laws covering copyrights and trademarks protect specific works and brands, not a person’s identity.

In the absence of stronger safeguards, stars like Taylor Swift and Lionel Richie have started registering trademarks on their voices, but it’s unclear how effective such efforts will be. Attorneys for artists have also used copyrights and old state likeness laws to combat fakes online, but those laws aren’t a perfect fit for addressing deepfakes.

NO FAKES would address that gap by allowing individuals to sue anyone who posts unauthorized “digital replicas” of their likeness, or the technology companies that enable their creation. It would come with a system of safe harbors that shield online platforms from such liability if they quickly remove such content, similar to the existing system for copyright takedowns. But the new bill would impose even tighter rules on removal: Platforms would be required to ensure that the same content is not quickly re-uploaded, a common gripe about the current copyright rules.

The new likeness power created by the bill would be a property right, which would extend beyond a person’s death and could be controlled by their heirs for decades. It can be licensed like any other piece of intellectual property, albeit with a 10-year cap on such licenses.

First introduced in 2024, NO FAKES has drawn criticism for the potential to harm free speech. Opponents have warned that it could be abused with frivolous takedowns or lawsuits against legal content, and that it would incentivize platforms to quickly remove even legitimate materials out of fear of liability. Key updates have been made in an effort to address such concerns, including explicit carve-outs for news coverage, biopics and criticism, and beefed up rules for restoring online content and punishing bad faith removals.

Over three iterations, NO FAKES has slowly built a broad coalition of supporters, including the three major music companies, the Recording Industry Association of America (RIAA), the Recording Academy and the National Music Publishers Association (NMPA); tech giants like OpenAI, Google and TikTok; music streaming platforms like Spotify; the movie industry via the Motion Picture Association; and major unions like SAG-AFTRA and AFL-CIO.

“Unchecked AI can ruin lives,” said SAG-AFTRA president Sean Astin a day before the vote. “Americans are demanding that the federal government take sensible action. The NO FAKES Act would establish a fundamental protection to control their own voice and likeness.”

But critics of the bill remain. Digital rights groups like Public Knowledge and the Electronic Frontier Foundation have continued to voice free speech concerns. And days before Thursday’s vote, the Entertainment Software Association, the lobby for the massive video game industry, urged the committee to reject NO FAKES in its current form.

“The bill makes no distinction between harmful deepfakes and legitimate digital replicas, such as those in video games, the group wrote. “[It] threatens to engender frivolous lawsuits by those who may, even by coincidence, resemble a game character, especially one of the thousands of background characters present in video games.”

Such concerns could still be addressed in future versions of the bill. Though Thursday’s vote was a key hurdle — most Congressional bills never make it out of committee — NO FAKES can and likely will be revised further as it moves toward a full vote.

In a statement, RIAA Chairman & CEO Mitch Glazier applauded Thursday’s vote: “An extraordinary cross-sector coalition… have come together to support these protections for Americans’ voice and likeness from exploitive digital deepfakes, and consumers agree: 92% worry about the impact of AI deepfakes on authenticity, society and culture. The NO FAKES Act answers the call.”

The Human Artistry Campaign, an advocacy group that has pushed for AI restrictions, voiced similar support: “Creativity is rooted in human experience — perspectives, faces and stories that connect us and move culture forward. As AI evolves, everyone deserves the right to control how their voice, likeness and identity are used.”

Daily newsletters straight to your inbox

Sign Up