On June 9, 2026, New York became the first U.S. state to require advertisers to disclose when a "performer" in a commercial is generated by AI. The law was written in direct response to brands quietly deploying synthetic humans and hoping nobody would ask. Weeks earlier, The Guardian had traced an Instagram ad for the photo app Once back to a crying bride who does not exist. Shein and Guess had already run their own laps through the same backlash cycle. The law didn't invent a problem. It caught up to one the market had already priced wrong.
Why a fake human costs more trust than a bad ad ever could
A bad ad is a creative failure; a fake human is a broken social contract. When 62% of consumers say they trust a brand less after discovering undisclosed AI in its advertising, they aren't reacting to production quality. Once's crying bride was technically convincing. They're reacting to the deception itself: something that asked for empathy and para-social trust turned out to be a puppet. A testimonial is processed through a different cognitive channel than an ad is. Reclassifying "witness" as "asset" after the fact doesn't just cost that campaign. It taxes every claim the brand makes afterward. The base rates back this up: 81% of consumers say they've stopped supporting a brand once it stopped feeling genuine, and 97% count authenticity among the reasons they pick one brand over another. Those aren't soft sentiment numbers. They're churn numbers wearing a survey costume.
New York's $1,000 fine is the cheapest part of this story
Brands reached for synthetic performers for the same reason they reached for programmatic buying: the unit cost was lower. No talent fee, no usage-rights renewal, no risk of a creator saying something off-brand at 2 a.m. That math never priced in the cost of discovery. The statute's penalties ($1,000 for a first violation, $5,000 after) are a rounding error next to what an undisclosed AI performer costs once caught. The FTC's ceiling for a knowing endorsement violation runs to $53,088 per post; a single campaign with a dozen placements can theoretically expose a brand to seven figures before a lawyer opens the file. But regulatory exposure is the smaller number. GlobeScan's 2026 Radar survey, sampling 27,000 consumers across 30 countries, recorded boycott incidents up 34% year-over-year, the highest rate since the study began tracking it in 2014. Getting caught doesn't just draw a fine; it feeds a distribution mechanism that was already primed to punish you.
This isn't a marketing problem. It's the same distrust wave hitting every institution
PwC's 2026 Voice of the Consumer survey found 82% of B2C executives believe consumers trust their brand highly, while only 49% of consumers agree. That is a 33-point gap between what marketers think they've earned and what they've actually banked. McKinsey's 2026 State of the Consumer survey found Gen Z reports lower trust than baby boomers across nearly every research channel despite living on social platforms all day. The heaviest users are also the most skeptical. That gap is why "de-influencing" content, where creators openly warn audiences off a purchase, now reads as more credible than a polished endorsement: it signals a human choosing candor over commission. NIQ's June 2026 outlook backs the same pattern from a different angle: functional quality and plain responsiveness now outweigh diversity or sustainability messaging as the biggest trust driver in every market surveyed. The bar was never perfection. It was competence and candor. Brands reaching for synthetic performers to manufacture relatability picked the exact moment culture stopped rewarding anything manufactured. The law is downstream of the sentiment, not the other way around.
What this means for your next influencer brief
The fine was never the risk. The reclassification is.
- Disclose before you're forced to. Label synthetic content clearly, and let the honesty do work a bigger budget can't buy.
- Audit every testimonial in-flight, now. If you can't produce the release form or the real person behind a "customer voice," assume a regulator or a journalist eventually will.
- Rebuild the real-creator line item. The messiness that made human creators harder to brief is exactly what now reads as trustworthy.
- Plan for the law to travel. GDPR started in one jurisdiction and became the default; budget for compliance now, not after a second state (or a second market) passes its own version.
New York priced the fake human at $1,000. The audience had already priced it at 62%.
The brands that get ahead of this aren't the ones with the cleanest legal defense. They're the ones who never needed one.
Want this kind of thinking on your brand?
We build brand strategy, AI content and performance for the AI era.
START A PROJECT →