How new AI disclosure laws are reshaping fashion advertising

For fashion brands, the question of whether to label AI-generated imagery has stopped being a matter of principle and become a matter of law.

In the U.S., New York’s Synthetic Performer Disclosure Law took effect on June 9, requiring any advertisement featuring a “synthetic performer” to include a clear and conspicuous disclosure within the piece. The law applies to any company whose ads reach New York consumers, regardless of where the advertiser is based. On August 2, the EU Artificial Intelligence Act will take effect, reinforcing disclosure obligations regarding companies' use of AI-generated or manipulated image, audio, or video content.

Evidence of the impact is already visible in production. Caimera, the AI imagery platform that works with more than 200 enterprise brands, told FashionUnited the regulation is reshaping the brief before it reaches the disclosure line. “There are some clients who have decided to switch the type of AI content, for example flat lay or ghost shots without models, or use it for design teams,” Caimera's co-founder Kirti Poonia said.

The specifics of each law explain why. New York’s statute, signed by Governor Kathy Hochul in December 2025 and billed as first-in-the-nation, targets the advertiser that produces or creates the ad, applies where the advertiser has actual knowledge that a synthetic performer is used, and carries civil penalties of 1,000 dollars for a first violation and 5,000 dollars for each subsequent one. Audio-only ads, promotional material for expressive works such as films and video games, and AI used solely for language translation are exempt, and publishers that merely disseminate a non-compliant ad are shielded.

The EU rules cut wider, with provisions applied even without intent to deceive, including content that looks or sounds like a real person must be labeled even if no deception was intended and even if no real individual is depicted. According to the regulation, disclosure must reach the viewer clearly and at first exposure, and cannot be buried in terms and conditions or left to machine-readable metadata alone.

For fashion, the exposure is concentrated in on-model campaign imagery, one of the areas brands have been most eager to automate. Consumers want transparency. Caimera’s 2026 survey of 502 U.S. consumers found that 85 percent could not reliably tell AI-generated images from real ones, while 75 percent still believe AI imagery should be disclosed, and when two brands both use AI, 79 percent said they would trust the one that labels it.

The consumer response

Last December, Valentino published an image created with the help of AI to promote one of its handbags on social media. The backlash came right after.

Dr. Rebecca Swift, senior vice president of creative at Getty Images, told the BBC at the time that the negative reaction suggested many see AI content as "less valuable" than human creations. "While people are excited by AI-generated content for personal use, they hold brands to a higher standard, especially expensive brands," she said. "Even full transparency about AI use wasn't enough to win them over."

Diesel, Gucci, Collina Strada, Baggu, Selkie, Mango, H&M, Zalando, Guess and Levi’s are just a few of the brands that have also faced public criticism after using generative AI models or imagery in their campaigns.

Questions about IP, exploitation and environmental impact have also been previously pointed out online by consumers as reasons for concerns about the use of AI in fashion. Those concerns aren't abstract: a New York model sued Rainbow Shops this spring over AI images generated from an expired contract; the state's new Fashion Workers Act now requires consent for models' digital replicas, and image generation remains one of the most carbon-intensive AI tasks, according to The Sustainable Agency, undercutting the sustainability case brands often make for it.

The rules are unlikely to push brands away from AI altogether

According to Caimera’s report, the AI visualisation during the design phase can cut sampling costs by roughly 45 percent and marketing production costs by as much as 80 percent, while compressing a six-month concept-to-launch cycle by up to four months. “AI helps run the whole workflow more efficiently and sustainably; there is at times no sampling as sketches turn into CADs and then into On-model images using AI,” Poonia said.

It is also reshaping roles in the industry. As AI moves from concepting into production, Poonia expects the impact to fall unevenly across a marketing team. "Marketing has always had two types of people, the thinkers/ideators and the executioners," she said. "In the future, the first set will have the same skills and the same jobs; the second set will have to move to embracing AI skills, just as analog photographers had to learn digital photography."