Where the law stands on AI generated performers in advertising: the New York synthetic performer rules, the NO FAKES Act, state right-of-publicity expansion, and the practical checks before you run a generated face.
HighQualityUGC Team
Editorial
We run UGC ad tests daily and publish what holds up: real credit costs, real hook rates, no vendor fluff.
The question has two halves that get confused, and separating them makes the answer much clearer.
The first half is whose face is it. Using a synthetic person who resembles nobody in particular is a different legal situation from generating something that looks like an identifiable human being. The second half is do you have to say so. That is a disclosure question and it varies by jurisdiction.
Almost every real problem in this area is the first half. Almost every new regulation is about the second.
Right of publicity is the body of law that gives a person control over the commercial use of their name, image, voice and likeness. It is mostly state law in the United States, and several states, including New York, California, Illinois and Tennessee, have enacted or expanded these protections specifically to cover AI generated replicas.
What this means in practice:
At the federal level the NO FAKES Act, advanced unanimously out of the Senate Judiciary Committee on 18 June 2026, would create a federal property right over voice and visual likeness in digital replicas. It has not become law. Its movement matters because it signals where this is heading, but the operative rules today remain state law.
Effective 9 June 2026, New York Senate Bill S8420A requires advertisers to conspicuously disclose when an advertisement contains a synthetic performer, defined as a digitally created asset generated with AI or software algorithms intended to give the impression of a human performer who is not recognisable as any identifiable natural person.
Read the definition carefully, because it is doing something specific. The New York rule targets exactly the low risk case from the previous section: a synthetic human who is nobody. If your generated actor is realistic enough that a viewer might take them for a real person, the disclosure obligation attaches.
So the two halves interact. Making the face unrecognisable as any real person solves the likeness problem and triggers the disclosure requirement.
The common objection is that a disclosure label will hurt performance. The available evidence points the other way, and the reasoning is intuitive: viewers are already suspicious, and confirming what they suspected converts better than being caught.
The worse outcome is discovery. A comment section that works out your spokesperson is not real, on an ad with no label, is a trust problem that outlives the campaign.
The rules are not uniform and they are moving. The EU AI Act includes transparency obligations for synthetic content, several countries have advertising standards guidance on AI generated imagery, and platform policies add another layer on top of law: Meta, TikTok and others have their own AI content labelling requirements that apply regardless of jurisdiction.
If you run internationally, the practical approach is to disclose everywhere rather than maintain a per-market matrix. One version of the ad, labelled, is cheaper than five versions and a compliance review.
Being honest about the boundaries:
None of this blocks ordinary use of a synthetic actor who is nobody in particular, disclosed where required. All of it argues for keeping records and avoiding the temptation to generate something that looks like a specific real person.
Generally yes, provided the actor is not a recognisable real person used without permission, and provided you disclose where disclosure is required. New York's synthetic performer law, effective 9 June 2026, requires conspicuous disclosure of AI generated performers in advertising, with penalties of $1,000 for a first offence and $5,000 subsequently.
No, not without permission. Right of publicity law in states including New York, California, Illinois and Tennessee has been expanded to cover AI generated replicas of a person's likeness and voice, and that a model produced the image is not a defence. Accidental resemblance is a real risk worth checking for.
A federal bill that would create a property right over an individual's voice and visual likeness in digital replicas, holding producers and distributors of unauthorised replicas liable. It advanced unanimously out of the Senate Judiciary Committee on 18 June 2026 but has not become law, so state right-of-publicity rules remain the operative law today.
In New York, yes, when the ad contains a synthetic performer intended to give the impression of a human. Other jurisdictions and platform policies add their own requirements. If you advertise across markets, disclosing everywhere is usually cheaper than maintaining a per-market matrix.
The evidence points the other way. Viewers are already suspicious of polished video, and confirming what they suspected tends to build trust. The costly outcome is discovery: a comment section working out that an unlabelled spokesperson is synthetic damages the brand well past the campaign.
