Amazon now requires sellers to label AI-generated people in product images

A new New York State law pushed Amazon to act. Here is what it means for shoppers scrolling through listings and for the sellers who make their living on the platform.

AI2Day NewsdeskAI-assistedPublished Updated Editor: Lee Brown3 min read
Illustration: a darkened laptop screen displaying a generic seized-domain notice in a modern living room
Illustration made with AI. Not a photograph of the events described.
Share

Key points

  • Amazon told third-party sellers on Wednesday that they must tag product images and videos containing photorealistic AI-generated people before uploading them.
  • New York State's law, which took effect last month, requires any business advertising in the state to disclose when a "synthetic performer" (a computer-generated image made to look like a real human) appears in place of a real actor.
  • Governor Kathy Hochul called it a "first-in-the-nation" law, and Amazon's policy change follows directly from it.
  • Outside sellers account for more than 60% of goods sold on Amazon's marketplace, so this rule touches a huge share of the listings shoppers browse every day.
  • No federal law yet requires AI-content disclosure in advertising; individual states are filling that gap.

If you've ever clicked an Amazon listing and wondered whether the model in the product photo was a real person, Amazon now has to tell you. Starting this week, sellers who use photorealistic AI-generated people in images or videos must attach specific metadata tags to that content before it goes live.

Amazon confirmed the change in an announcement sent to sellers Wednesday, first reported by CNBC Tech. The policy covers standard product photos and "A+ content," the richer videos and graphics brands add lower down on a listing page.

Why is Amazon doing this now?

New York State passed a law that took effect last month requiring companies to disclose when ads feature "synthetic performers," the legal term for AI-generated figures designed to look like real humans. Amazon sells to customers across the country, so complying with a state law effectively means changing platform-wide behaviour.

Governor Kathy Hochul's office warned that without disclosure, AI-generated synthetic performers "can undermine one's ability to accurately distill fact from fiction." If a "person" selling you a vitamin supplement never existed, you probably deserve to know that.

Amazon says it will add a visible indicator to affected listings. The company hasn't yet spelled out exactly what criteria it will apply when deciding when to show that label.

We first covered the New York AI law on 27 July 2026, and the platform-level pressure it's now producing is exactly what that piece flagged as the likely next step.

What does this mean for sellers?

Sellers need to act before uploading, not after. The requirement is to embed specific metadata keywords in image and video files. Amazon hasn't released its full technical guidance publicly, so sellers should watch their Seller Central dashboard for details.

A few things the new rule does NOT cover: content featuring fictional characters from TV or video games, and images of real people even if AI tools were used to touch them up.

Content type Label required?
Photorealistic AI-generated person Yes
Real person, AI-edited No
TV or game character No
Product-only image, no people No

More sellers have been turning to AI tools to generate listing images quickly and cheaply. That's a real business advantage. This rule doesn't stop them; it just adds a disclosure step when the output looks like a human being.

The broader pattern is worth watching. As we reported on 22 July 2026, Substack is now flagging AI-written newsletters before readers open them. Amazon's move suggests that disclosure is becoming the floor, not the ceiling, for platforms that want to stay ahead of state legislators.

Common questions

Does this affect shoppers outside New York?

Amazon has applied the policy across its US marketplace, so shoppers in every state will eventually see AI-content labels on qualifying listings.

Is there a national law covering this?

Not yet. No federal rule requires AI-content disclosure in advertising. California has moved separately to require watermarks inside AI-generated media files, but New York is the first state to specifically target synthetic performers in ads.

© 2026 AI2Day