THE PULSEBrand dealsPlatform payoutsCreator fundingUGCSocial commerceAI & creators
the business of creativity
Opinion

AI labelling laws are the best thing to happen to human creators

New York and the EU now require synthetic performers in ads to be labelled. Creators who see this as red tape are missing the point: a visible label turns "human" into a selling point.

By Editorial Board · October 30, 2026
Photo illustration: The Collab Report

For two years, human creators have watched AI avatars learn to hold a serum bottle, smile at the camera and say "honestly, this changed my skin." The pitch to brands was irresistible: the look of a creator, without the creator. This year, lawmakers in New York and Brussels quietly broke that pitch. Creators should be cheering, loudly.

What changed this year

Two rules matter most.

In New York, Governor Kathy Hochul signed S.8420-A/A.8887-B on December 11, 2025, and it took effect on June 9, 2026. According to analyses by law firms Cooley and Hunton, it requires advertisers to conspicuously disclose when an ad includes a "synthetic performer." The law defines that as a digitally created asset, made with generative AI or other software, that gives the impression of a human performance by someone who isn't an identifiable real performer.

The duty falls on those who produce or create the ad for a commercial purpose and know a synthetic performer is in it. Media outlets and platforms that merely carry the ad are exempt. So are audio-only ads, ads for films, TV shows and video games that use a synthetic performer consistently with the work itself, and AI used only to translate a human performer's words. Penalties are $1,000 for a first violation and $5,000 for each one after that.

In the EU, the AI Act's labelling obligations under Article 50(4) took effect on 2 August 2026. Law firm Lewis Silkin's analysis says a realistic AI celebrity or synthetic influencer promoting a product will likely count as a deepfake that must be labelled. The label has to be visible to people; invisible watermarks alone won't do. Fines can reach €15 million or 3% of global turnover, and the rules can reach advertisers outside the EU who knowingly distribute content there.

Why this helps people

The case for synthetic UGC was never that it was better. It was that it was cheaper and indistinguishable. These laws remove the second half of that case in two of the world's most important ad markets.

Once the AI performer must carry a label, the ad is no longer "a real person who loves this serum." It's "a generated person scripted to love this serum." For some formats, that won't matter. Explainers, product demos and obvious fantasy work fine with a label attached. But the core value of creator marketing has always been perceived authenticity: a real person, with a real audience, vouching for something. A mandatory label strips that out by law. Human creators get a competitive advantage they didn't have to lobby for.

The evidence that synthetic influencers outperform humans was thin to begin with. CreatorDB, a creator data company, notes in its explainer on AI influencers that there is still no data comparing their effectiveness with human creators, and that audiences may find humans more relatable for some products. The marketing claims ran well ahead of the proof.

Meanwhile, brands keep saying they want authenticity. In CreatorIQ's 2026 creator survey, the brands questioned ranked follower count last among their partnership criteria and talked up fit and suitability. A labelled synthetic performer is a hard sell on fit with an audience that came for a person.

Note what New York did not do. It didn't ban synthetic performers, and it didn't prescribe the wording or form of the disclosure. Hunton points out that the statute leaves both open. That restraint is a strength. Brands remain free to use AI where it genuinely helps, which makes the law harder to attack as anti-innovation. All it removes is the option of passing a generated person off as a real one.

The EU went further in one respect: its label has to be perceivable by people, not buried in metadata. Together, the two regimes cover a huge share of the global brands that buy creator content. A multinational advertiser running one campaign across markets has every reason to apply the stricter standard everywhere, rather than maintain separate labelled and unlabelled versions.

What creators should do with this

  • Make "human-made" explicit. If AI performers must carry a label, a human creator's work implicitly carries the opposite one. Say so in your media kit and your pitch.
  • Put AI clauses in contracts. Ask whether a brand plans to clone your likeness or voice, or generate "versions" of you. Undisclosed synthetic versions are now a legal risk for the advertiser in New York and the EU, and that gives you leverage.
  • Use AI as a tool, not a replacement. CreatorIQ found that 72% of creators have used AI tools, mostly for brainstorming and captions, and only a small minority for strategy or automation. That's fine. An AI-assisted caption doesn't make you a synthetic performer.
  • Know the gaps. New York's law doesn't cover audio-only ads, and it targets ad creators, not platforms. Voice talent and podcast hosts shouldn't assume the label protects them.

The strongest counterargument

The serious rebuttal is that labels fail because people stop seeing them. Think of cookie banners, or the way "#ad" became background noise. If audiences scroll straight past the AI tag, synthetic UGC keeps its cost advantage and loses nothing. Penalties of $1,000 to $5,000 are also small for a large advertiser; some will treat them as a cost of doing business.

That's a real risk, and it's why creators shouldn't treat the label as a finished victory. But compliance doesn't only cost the fine. It costs legal review, approval workflows and the reputational damage of being caught passing off a synthetic person as real. In the EU, the ceiling is far higher. For many brands selling into New York or Europe, a human creator will simply be the lower-risk option.

There is also the question of enforcement, which is still untested in New York. But laws shape behaviour before the first case is brought. Agency lawyers are already writing disclosure into their approval checklists, and every checklist that asks "is this person real?" is a small win for people who are.

Labels won't save every UGC job. What they do is put a price on pretending. For creators, it's the first time the law has made being real a commercial advantage. Don't waste it.