New York’s Synthetic Performer Law: What Brands Need to Know

AI has officially entered its fine print era.

AI is making it faster, easier, and cheaper to create almost anything: images, voices, videos, spokespeople, customers, influencers, even fully synthetic human beings.

As the tools get more powerful, audiences are beginning to wonder …

Who is real and who isn’t?

Who gave permission to be featured?

And does the audience deserve to know?

In New York, the answer to that last question is now very clearly: Yes.

On December 11, 2025, New York signed a new law requiring advertisements to disclose when they include a “synthetic performer.” The law took effect June 9, 2026, and applies to ads appearing in New York State. In plain English: If an ad includes a digitally created, human-like performer, then the ad needs to clearly say so.

This is not just another AI regulation to toss into the legal junk drawer next to privacy policies, cookie banners, and the unread terms and conditions we all pretend we’ve carefully reviewed.

This one matters because it sits directly at the intersection of creativity, commerce, technology, labor, likeness, and trust.

So, let’s unpack what brands need to know.

The law, without the legal fog

The New York law requires advertisements to “conspicuously disclose” when a synthetic performer appears in the ad. The law also includes penalties: $1,000 for a first violation and $5,000 for subsequent violations.

The important part for brands and agencies is understanding what actually triggers the requirement.

This is not about whether AI was used somewhere in the creative process. AI might have helped write a headline, generate a background, resize an image, brainstorm concepts, or build a moodboard. That is not the core issue here.

The key question is much more specific:

Does the ad show an AI-generated or software-created person who looks human but is not actually a real, identifiable person?

If yes, then disclosure is required. If no, then this specific law may not apply. If the answer is “hmm, not totally sure,” that is probably the moment to pause before hitting publish. Internal guidance should treat synthetic human performers as a specific high-risk advertising category requiring review before release.

Think synthetic spokespeople. AI-generated models. Realistic avatars. Fake customers. Human-looking “employees” who have never filled out a W-2, joined a Zoom call, or eaten the last bagel in the breakroom.

If they look like people, perform like people, and appear in an ad, but they are not real people, New York wants audiences to know.

What this means for brands

The immediate takeaway is compliance. Brands running ads in New York need a process for identifying synthetic performers, approving disclosure language, and making sure the disclosure appears clearly on the ad itself.

Key things brands should know

First: The disclosure needs to be clear and visible. The law uses the word “conspicuous,” which is doing a lot of work here. Until more enforcement guidance develops, brands should assume the disclosure needs to be easy for a reasonable person to notice and understand at the time they see the ad.
Second: The disclosure should live on the ad itself. Not hidden in metadata. Not buried in a caption. Not tucked into a campaign landing page no one will visit. Not whispered softly in the legal notes while the synthetic performer smiles brightly in the hero frame.
Third: Video needs special attention. If a synthetic performer appears on screen, the safer approach is to keep the disclosure visible while that performer appears. A blink-and-you-missed-it disclaimer at the beginning or end may not do the job.
Fourth: Every version matters. The :30, the :15, the :06, the social cutdown, the paid placement, the organic post, the reel, the display unit, the OOH board, the CTV spot—all of it needs review if the synthetic performer appears.
Fifth: “It’s obviously AI” may not be enough. Creative teams live close to the work. We know what was generated, composited, rendered, retouched, mocked up, prompted, polished, and Frankensteined into existence at 11:47 p.m. on a Tuesday. Audiences don’t. And the law is not written for the people who made the thing. It is written for the people seeing it.
Sixth: Brands should document the decision. Who reviewed it? What disclosure language was used? Where did it appear? Which versions were checked and by whom? Documentation may not be glamorous, but neither is scrambling after launch to figure out which cutdown went where and whether the disclosure survived the resize.

What this means for audiences

Advertising has always involved performance. Actors play customers. Models embody lifestyles. Spokespeople simplify complex products into approachable human moments. None of that is new. But synthetic performers add a new layer because the “person” on screen may not be a person at all.

That matters.

It matters for talent, whose likeness, labor, craft, and opportunity are all being reshaped by generative tools. It matters for representation, because synthetic humans can easily become frictionless composites of what a brand thinks people should look like. It matters for audiences, because believability is part of persuasion. And it matters for brands, because the quickest way to damage trust is to make people feel tricked.

When a brand creates a person who never existed to sell something to people who do, there is a trust gap. Disclosure helps close it.

Not because audiences are fragile. They’re not. People understand that advertising is constructed and has more than a little theater involved.

Transparency gives audiences the dignity of context. It says: This is synthetic, and we trust you enough to tell you. That small act can make the difference between a brand using new tools responsibly and a brand using new tools to blur reality for convenience.

Don’t fret: transparency does not have to hurt performance

Here’s the part brands should find reassuring: disclosure does not appear to be the creative death sentence some may fear.

A recent Marketing Dive report on research from MediaScience, conducted with MediaPet and the Ehrenberg-Bass Institute for Marketing Science at Adelaide University, found that AI disclosure labels did not meaningfully hurt video ad performance. The study showed minimal negative impact on measures like brand recognition, ad sentiment, and brand attitude.

Consider: That’s a big deal because it suggests that the industry may be worrying about the wrong thing. The label is not the enemy. Bad creative is.

A disclosure will not save weak work. But it also does not automatically ruin strong work. In fact, the better opportunity is to make transparency part of the creative standard: a small act of honesty in a marketplace where audiences are increasingly unsure what is real, what is generated, and who is behind the message.

AI will continue to change how advertising gets made. Consider: That’s exciting. We should explore it. We should experiment. We should use the tools.

But we should not confuse possibility with permission.

So yes, disclose the synthetic performer. Make the label readable. Keep it with the asset. Build it into the workflow. Ask the uncomfortable questions early.

Because AI can fake a face. It can’t fake trust.