If your ads use an AI-generated person, New York now requires you to say so
This is the first law of its kind in the United States, it took effect last month, and it applies to any company whose ads reach a New York audience — including a roofing company in Texas running Meta ads with an AI-generated homeowner.
What the law requires
New York Governor Kathy Hochul signed S.8420-A / A.8887-B on 11 December 2025. It took effect 9 June 2026.
The requirement: when a visual or audiovisual advertisement distributed to a New York audience features an AI-generated synthetic performer — a fabricated human likeness that does not depict a real person — the ad must conspicuously disclose that a synthetic performer appears in it.
The duty attaches to any person who "produces or creates an advertisement" for a commercial purpose, in any medium, where that person has actual knowledge that a synthetic performer was used.
New York's announcement describes it as the first law in the nation to require this.
Why this applies to you even if you are not in New York
The trigger is the audience, not your address. The law reaches any company whose ads reach consumers in New York, regardless of where the advertiser is headquartered.
Consider how ordinary that is for a contractor. You run Meta ads. Meta's targeting is imperfect and your creative gets served beyond your intended radius. You used a stock-style AI-generated "homeowner" in the thumbnail because it was cheaper than a shoot. You now have a disclosure obligation you have not thought about.
The penalties are civil: $1,000 for a first violation and $5,000 per subsequent violation. Not existential. Also not the kind of letter you want to explain to a franchisor or an insurer.
What is exempt
Three carve-outs worth knowing, because they narrow this more than the headlines suggest.
- Audio-only advertisements. A radio or podcast spot with a synthetic voice is outside this law's scope.
- AI used solely for language translation. Localising your ad does not trigger it.
- Advertisements for expressive works — films, TV and similar — where the synthetic performer's use in the ad is consistent with its use in the work itself.
Note what is not exempt: AI-generated backgrounds, property imagery, before and after composites, or a synthetic voice in a video. The statute is about a fabricated human likeness in visual or audiovisual advertising, and that is a category most contractor creative can fall into by accident.
The practical checklist
The law does not specify the form of the disclosure — no required language, placement or size. That is genuinely unhelpful, and it means erring toward clarity.
- Inventory your creative. Which live assets contain a human figure that was generated rather than photographed? If you cannot answer, that is the first problem.
- Ask your agency in writing. "Actual knowledge" is the standard. An agency that quietly used generative imagery has handed you an obligation without telling you.
- Decide a house rule. Either no synthetic human likenesses in paid creative, or a standard disclosure applied every time. Both are defensible; ad-hoc is not.
- Keep the provenance. Which tool, which prompt, which asset. If someone asks whether a face is real, you want a file, not a memory.
Our own position, since it is fair to ask: there is not a single generated human likeness anywhere in the creative on this site. The photography in this section is generated, and it contains no people — that is a deliberate line, not an accident.