AI Deepfakes and Your Right of Publicity

Generative AI has made it trivial to fabricate a person saying or doing something they never said or did. For founders, creators, and athletes whose identity is the asset, an AI deepfake is not a novelty — it’s theft of the thing they’ve spent years building. This guide takes the enforcement view: how the right of publicity applies when someone clones your face or voice, which new laws give you teeth, and exactly what to do about it.

What is the right of publicity, and does it cover AI deepfakes?

The right of publicity is a state-law right that lets you control — and profit from — the commercial use of your identity. That identity classically includes your name, image, and likeness (NIL), and courts have long extended it to distinctive attributes like a signature phrase, a look, and, importantly, your voice. It exists to stop others from cashing in on the persona you built.

Crucially, the right doesn’t require that a real photo or recording of you was used. Courts recognized “identity” theft long before AI: in Midler v. Ford Motor Co. (9th Cir. 1988) a sound-alike singer imitating Bette Midler was enough, and in White v. Samsung (9th Cir. 1992) a robot dressed like Vanna White evoked her identity. An AI deepfake is a more perfect version of the same wrong — it evokes and appropriates you to sell something or trade on your fame. If the deepfake is used commercially without consent, it fits squarely within traditional right-of-publicity doctrine.

For the foundations — how NIL rights work, who owns them, and how licensing operates — start with our right of publicity and NIL guide.

Which states have new AI “digital replica” laws?

Because the right of publicity is a patchwork of state laws (there is no general federal statute), a wave of 2024–2025 legislation has updated it specifically for AI. The two most important:

• Tennessee’s ELVIS Act — the Ensuring Likeness, Voice, and Image Security Act, effective July 1, 2024. It amended Tennessee’s Personal Rights Protection Act to add voice as a protected element alongside name, photograph, and likeness, and to explicitly reach AI-generated simulations. Its signature move: it imposes liability not just on the person who publishes an unauthorized replica, but on anyone who makes available a tool or service whose primary purpose is creating one. That aims at the deepfake pipeline itself.
• California’s AB 1836 and AB 2602 — both effective January 1, 2025. AB 1836 amended Civil Code § 3344.1 to require consent from a deceased personality’s estate before using a digital replica of them in most expressive works. AB 2602 (codified in the Labor Code) voids contract clauses that let a studio use a digital replica of a performer’s voice or likeness in place of their real work unless the use is specifically described and the performer had legal or union representation.

Other states have moved as well. Illinois amended its Right of Publicity Act to address digital replicas, New York’s Civil Rights Law now covers digital replicas of deceased performers and non-consensual synthetic media, and numerous states have added deepfake provisions aimed at political advertising or non-consensual intimate imagery. The result is a genuine patchwork — a deepfake that’s clearly actionable in Tennessee or California may fall into a gap in a state that hasn’t updated its statute.

The practical takeaway: the applicable rules depend heavily on where you are domiciled and where the infringer operates, because many of these rights are tied to the plaintiff’s residence at the relevant time. That choice-of-law question is often the first thing a lawyer will analyze. For the deep dive on how these statutes fit together, see the NO FAKES Act and digital-replica laws.

What is the NO FAKES Act, and is it law yet?

The NO FAKES Act — Nurture Originals, Foster Art, and Keep Entertainment Safe Act — is a proposed federal bill, reintroduced in Congress in 2024 and again in 2025. As of 2026 it has not been enacted, so it does not yet give you any rights. But it’s worth understanding because it would fill the biggest gap in current law: the absence of a uniform national right.

If passed, the NO FAKES Act would:

• Create a federal, IP-style digital-replica right covering both voice and visual likeness, applicable to living and (for a term of years) deceased individuals;
• Make it unlawful to produce or distribute an unauthorized digital replica, and reach platforms that host them;
• Establish a notice-and-takedown system for online services modeled on the DMCA’s Section 512 safe harbor, so a platform that promptly removes a flagged replica gains protection from liability;
• Include carve-outs for news, commentary, parody, and other First Amendment–protected uses.

Because it isn’t law, your enforcement toolkit today still runs on state right-of-publicity claims plus the new state digital-replica statutes above. We track the bill’s status and mechanics in our dedicated NO FAKES Act guide.

Is an AI voice clone treated the same as a face deepfake?

Legally, yes — voice is protected identity, and AI voice cloning is one of the fastest-growing categories of publicity abuse. The Midler line of cases established decades ago that your voice is part of your protectable persona, and the ELVIS Act now names voice explicitly. A cloned voice used to fake an endorsement, a scam robocall, or a “new song” you never recorded is an appropriation of your identity just as much as a fabricated video is.

Voice cases do carry practical wrinkles: proving the clone is recognizably you (rather than a generic synthetic voice) is a fact question, and the analysis can turn on how distinctive your voice is and how the clone was marketed. Where the cloned audio was explicitly promoted as “you” — or trained openly on your recordings — that recognizability is much easier to establish. Voice-clone disputes also frequently overlap with breach-of-contract and fraud theories, especially when a performer or narrator licensed limited use and the buyer generated far more than the deal allowed. We cover the specifics — training data, detection, and remedies unique to audio — in AI voice cloning and deepfakes.

How is this different from a commercial ad using my likeness?

The core wrong is the same — unauthorized commercial use of your identity — but AI deepfakes raise problems ordinary advertising cases don’t:

• No real source material. A traditional ad misuses an actual photo or clip. A deepfake fabricates one, so there’s no “original” to trace, and the defendant may claim the output is “synthetic” and therefore not really you. Digital-replica statutes exist precisely to close that argument.
• Scale and speed. One model can generate thousands of variations across platforms overnight, making whack-a-mole enforcement harder.
• Blended harms. A deepfake can be simultaneously a publicity violation (commercial appropriation), defamation (false statements attributed to you), and — in intimate contexts — a violation of specific non-consensual-imagery laws. Plead them together where the facts fit.

If your concern is a straightforward paid endorsement or ad that used your face or voice, the analysis and remedies are laid out in using someone’s likeness in advertising. Deepfakes borrow that framework and add the digital-replica layer on top.

What should I do first if an AI deepfake targets me?

Speed and evidence win these cases. In the first 48 hours:

1. Preserve everything. Screenshot the deepfake, capture the URL, note the posting account, timestamps, and any view/engagement counts. Save the file itself. Content disappears — your record shouldn’t.
2. File platform takedowns. Nearly every major platform now bans non-consensual synthetic media, impersonation, and manipulated media. Use the platform’s reporting flow first; it’s the fastest path to removal and often resolves the problem without litigation.
3. Send a cease-and-desist. A letter to the poster (and, where appropriate, the platform or the tool provider) demanding removal and preservation of evidence adds legal weight and starts the paper trail. Ground it in your state’s right-of-publicity statute and any applicable digital-replica law.
4. Identify the right defendants. That may include the individual creator, a company using the deepfake in marketing, and — under laws like the ELVIS Act — the provider of the tool built primarily to make replicas.
5. Call a litigation attorney. If the use is commercial, defamatory, or spreading, you may seek a temporary restraining order or preliminary injunction plus damages. As with any emergency relief, courts are skeptical of plaintiffs who waited.

What remedies and defenses come into play?

Your available remedies depend on the governing law, but typically include injunctive relief to force removal and stop further use, actual damages (including the fair market value of the misused identity and lost licensing revenue), the infringer’s profits, and, under some statutes, statutory damages and attorney’s fees. California’s § 3344, for example, provides statutory minimum damages plus profits and fees; the ELVIS Act likewise authorizes injunctions and damages. Willful conduct can raise the stakes.

Expect the defendant to raise the First Amendment. Right-of-publicity claims must coexist with free speech, so genuine news reporting, commentary, satire, parody, and transformative artistic works may be protected. Courts weigh whether the use is primarily commercial exploitation of your identity or expressive. A deepfake selling a product is squarely commercial; a clearly labeled parody sketch is a harder target. Anti-SLAPP statutes in states like California can also let a defendant seek early dismissal — another reason to build your factual record carefully before you sue.

To see how courts are applying these principles to real disputes, browse our right of publicity case archive.

The bottom line

An AI deepfake of your face or voice isn’t a new species of harm — it’s an old one, appropriation of your identity, supercharged by technology. The right of publicity already reaches it, and a fast-moving set of state digital-replica laws (Tennessee’s ELVIS Act, California’s AB 1836 and AB 2602) now name AI clones and voices explicitly, while the proposed federal NO FAKES Act would add a nationwide right if it passes. The playbook when you’re targeted is consistent: preserve evidence immediately, use platform takedowns first, escalate with a cease-and-desist grounded in the right statute, and bring suit for an injunction and damages when the use is commercial or defamatory. The faster you move, the more the law can do for you.

This guide is general education, not legal advice, and does not create an attorney-client relationship. Right-of-publicity and digital-replica laws vary sharply by state and are changing quickly — consult an attorney licensed in your jurisdiction before acting.