A New York disclosure rule is now changing how products look on the world's largest marketplace.
Amazon notified third-party sellers this week that any listing photo, video, or A+ content featuring a photorealistic AI-generated person must carry a disclosure label. The policy follows New York State legislation requiring companies to disclose when a synthetic performer replaces a human actor in advertising.
The demand on sellers is specific: inventory existing AI-generated assets, apply metadata tags to affected content, and bring listings into compliance. Amazon's marketplace hosts millions of third-party sellers, many of whom adopted AI image tools to cut the cost of product photography. That adoption now carries a compliance obligation that did not exist a year ago.
What the New York law requires
New York's synthetic-performer law targets a labor and consumer-transparency concern. When an AI-generated human likeness stands in for a real model or actor in commercial content, the law requires that fact to be disclosed. The statute does not ban the practice, it mandates acknowledgment of it.
Amazon's policy maps directly onto that requirement. The company still allows AI-generated people in product imagery. Sellers who use them must label them.
The labeling requirement creates an administrative burden, auditing existing content, applying tags, and maintaining compliance as new listings go live, without eliminating the cost savings that made AI imagery attractive. A blanket ban would instead have forced sellers to choose between AI tools and marketplace access.
The scope beyond New York
Amazon's enforcement is marketplace-wide, not limited to New York buyers or sellers. A seller based in Texas listing a product visible to shoppers in all fifty states must comply. Building state-specific content rules into a single global marketplace is technically and operationally impractical, so the most restrictive applicable standard becomes the floor for everyone.
This pattern has appeared before, most visibly in how California's privacy rules shaped data-handling practices far beyond the state's borders. New York's synthetic-performer law may follow a similar path, with Amazon's enforcement extending its reach nationally.
What sellers face now
The compliance task is heavy for sellers generating product images at scale. Identifying which images contain AI-generated people, as opposed to AI-modified backgrounds, AI-adjusted lighting, or other AI processing the policy does not appear to cover, requires either manual review or tooling that can make that distinction reliably.
Amazon has specified that sellers must apply metadata tags, which suggests the disclosure is embedded in the listing infrastructure rather than shown as on-page text to shoppers. How that tag surfaces to consumers, and whether it appears at all in the browsing experience, is not detailed in the available sourcing.
The policy went into effect with seller notifications issued around July 24. Sellers who have not yet audited their AI-generated content are already operating under the new requirement.
Get weekly leaderboard changes and AI video model launches delivered to your inbox.