The NO FAKES Act is a federal bill that would create a property right in a person's voice and likeness, enforceable against AI companies, platforms, and individual deepfake creators. It is the closest the US has ever come to a federal right of publicity. For creators whose face and voice are the entire commercial asset, the bill matters more than almost any other federal proposal currently in motion. Here is where it stands in June 2026, what it would do, and what creators should do now.
Background and current status
The NO FAKES Act was first introduced in 2023 and was reintroduced in updated form on 20 May 2026, with cosponsors from both parties. It cleared the Senate Judiciary Committee in June 2026, which is the furthest it has ever moved. A House companion bill is in committee. The current bill is more refined than the original draft, having absorbed feedback from the music industry (SoundExchange, RIAA), film industry (SAG-AFTRA, MPAA), platform operators (NetChoice, CCIA), and civil-liberties groups (the EFF).
The realistic outlook is that a vote in the Senate happens in late 2026 or early 2027, with passage uncertain but plausible. Even if the current bill stalls, the underlying policy direction is established and a successor bill in the next Congress is likely.
What "digital replica" means under the bill
The bill defines a digital replica as "a newly-created, computer-generated, electronic representation of the image, voice, or visual likeness of an individual that is so realistic that a reasonable person would believe it is an authentic representation of that individual." That definition is broader than most state-level deepfake laws. It covers AI-generated images, audio clones, video deepfakes, and synthetic performances. It does not cover traditional impersonation (someone wearing your makeup style), look-alike casting, or honest editorial commentary that uses your likeness with attribution.
The right under the bill is a property right. That means it can be licensed, sold, inherited, and enforced. It is also separate from privacy law, meaning a creator does not need to argue distress or reputational harm to enforce it: unauthorized use of the digital replica is itself the wrong.
The platform liability piece
The most contested element of the bill is the platform liability provision. The current draft establishes statutory damages of up to $750,000 per work for platforms that host unauthorized digital replicas after receiving notice. The notice-and-takedown regime is modelled on the DMCA: a rightsholder sends a properly-formatted notice, the platform has a defined window to remove the content, and a counter-notice procedure allows the uploader to dispute.
Crucially, the bill carves out Section 230 for digital replicas. That means platforms can no longer rely on the standard "we are not responsible for user-generated content" defence when they have notice of an unauthorized replica and fail to act. This is the change that makes the bill bite.
Platform operators (NetChoice, CCIA) have argued the bill creates a compliance burden that small platforms cannot meet and that the $750K figure is disproportionate. Their pushback is the main reason the bill has moved slowly. The current draft includes carve-outs for small platforms below a defined revenue threshold and for educational, news, and parody use, but the platform lobby remains opposed.
How it differs from state laws
State right-of-publicity laws already exist in most US states, and several states (Tennessee, California, Illinois, New York) have specifically expanded their statutes to cover AI-generated likeness. The Tennessee ELVIS Act has been in force since July 2024 and covers voice specifically.
State laws have three structural limits the NO FAKES Act addresses. First, jurisdictional scope: enforcing a Tennessee law against a platform with no Tennessee presence is hard. Federal law solves this. Second, the patchwork: fifty different state regimes create compliance complexity that disincentivises platforms from acting at all. Federal preemption (which the current bill includes) simplifies the picture. Third, Section 230: state laws cannot reach platforms because Section 230 is federal law. Only Congress can carve it out. The bill does.
The state laws will continue to matter, particularly for criminal enforcement (most state laws criminalise deepfake distribution) and for the strongest jurisdictional reach (a Tennessee court has standing over a Tennessee plaintiff in a way a federal court does not always). Use them in parallel rather than as substitutes.
What this means for creators
If the bill passes substantially as drafted, three things change for creators.
First, you have a federal cause of action against anyone making or distributing an AI replica of you without consent. You can sue. You can recover damages. You can get injunctions. The right is yours.
Second, platforms hosting deepfakes (Civitai, HuggingFace, Telegram channels, leak aggregators) must remove on notice or face platform-level liability. The notice-and-takedown regime gives you a fast administrative tool, similar to DMCA but specifically for likeness.
Third, AI labs face a clearer compliance environment. Labs that train on unauthorized creator likeness or generate replicas without consent face statutory damages. This is the change that finally creates an economic incentive for AI labs to negotiate consent and licensing.
What creators should do now to prepare
The bill is not law yet. But the work of being ready for it is work that pays off whether the bill passes or not.
Document your likeness baseline. Save high-quality reference photos of your face. Save voice samples (you can use a podcast appearance, a clean voiceover, anything that gives you a defensible reference of what your real voice sounds like). Save your handles and known content. The point is to have evidence on hand of what is authentically you so that you can prove a replica is unauthorized.
Register your handles. If you have not already, document the platforms where you operate, the handles you use, and any associated trademarks. Future enforcement is faster when your identity is already on a register that the legal process can reference.
Track AI replicas as they emerge. Use reverse image search and watchlists to identify when a deepfake of you appears. Save the URL, the date, the uploader if known, the source platform. This evidence is the foundation of any future suit. For the practical side of tracking and removing deepfakes, see our 2026 deepfake threat report.
Watch the legislative process. The bill will likely be amended further before passage. Specific provisions (platform-size thresholds, damages caps, fair-use carve-outs) will change. Creators who follow the bill closely will know what the final law actually says, rather than relying on press coverage.
How Privly is positioning for it
The work Privly is doing now is designed to be useful whether or not NO FAKES passes, and to be ready if it does.
Every Privly customer has the option to register their face, voice, and content fingerprints with us as a free protection feature. Creators who opt in build a verifiable baseline now. If NO FAKES passes, those creators are immediately positioned to enforce.
Our worker stack already detects unauthorized use of registered creator likeness across leak sites, deepfake repositories, and AI model marketplaces. Today the response is DMCA. After NO FAKES, it can become a dual notice under both DMCA and the new digital-replica regime.
Privly is publishing positions on the bill and will continue to engage with the legislative process. Creators who want their voice heard can use Privly's network to communicate with senators and representatives.
Realistic timelines and outlook
Three scenarios are worth holding in mind.
If NO FAKES passes in 2026 substantially as drafted, the operational impact on platforms and AI labs starts in 2027 (most bills include 6-12 month implementation windows). Creator enforcement starts the day the law goes into force.
If NO FAKES stalls but a successor bill passes in 2027 or 2028, the same dynamic plays out 12-24 months later. The policy direction is clear; the timeline is uncertain.
If no federal likeness law passes in this decade, state laws will continue to expand and the patchwork will get more complex. Creators will still have rights; they will just be harder to enforce across jurisdictions.
In all three scenarios, the work of building evidence, documenting baseline, and tracking infringement matters now. The legal regime that processes that evidence will arrive eventually.
The bottom line
The NO FAKES Act is the most important federal bill currently in motion for creators whose face and voice are their commercial asset. It is not law yet, but the underlying policy direction is established and the bill is closer to passage than at any prior point. Creators who treat this as a real near-term possibility and build their evidence baseline now will be ready when the law arrives. Privly is building the registry and enforcement infrastructure to support that readiness. And if you're being targeted by a deepfake or non-consensual intimate image right now, the Privly NCII assistance program is free and doesn't require an account.
