# Amazon orders sellers to label AI-generated people in product images after New York law > Source: > Published: 2026-07-24 07:46:59+00:00 *Amazon told third-party sellers on July 23 that they must tag product images and A+ content featuring AI-generated people with specific metadata before uploading, or face fines starting at $1,000 per image for sales to New York consumers.* If you've been using AI tools to generate product photography for your Amazon listings, you now have a compliance problem. Amazon informed sellers Wednesday of a new labeling requirement for any image or video containing AI-generated human likenesses, responding directly to New York State's synthetic performer disclosure law that took effect June 9, 2026. The announcement, first reported by CNBC from a copy of Amazon's seller notice, directs sellers to tag affected content with specific metadata keywords before uploading. Amazon will then display a consumer-facing indicator on those listings. The law Amazon is responding to is narrow in its text but broad in its practical reach. New York's rule targets what it calls "synthetic performers": digitally created human likenesses generated through AI or other software that create the impression of a real person. Under the statute, failure to disclose carries a $1,000 penalty for a first violation and $5,000 for each subsequent image sold to New York consumers. Those numbers don't sound punishing in isolation, but a seller with dozens of listings featuring AI-generated models, each one a separate violation, can reach serious liability fast. Amazon's announcement carves out some exceptions. The requirement doesn't apply to content featuring TV, video game, or movie characters, nor to images of real people who've been altered using AI. The target is fully synthetic human figures standing in for human models, the kind of AI product photography that has exploded across fashion, beauty, and lifestyle categories on the platform over the past two years. Here's the thing about a platform-level policy triggered by a single state law: it doesn't stay in that state. Amazon hasn't built a mechanism to show one version of a listing to New York shoppers and another to everyone else. That's the problem. So the practical effect of its July 23 announcement is that every seller on the platform, wherever they're based, wherever their customers shop, now operates under New York's disclosure standard. A boutique apparel brand in Austin using AI-generated models for its Amazon listings faces the same labeling obligation as a seller in Manhattan. The geography of the law is New York. The geography of enforcement is global. That dynamic, a single state law being effectively federalized by platform compliance, is exactly how California's privacy rules spread across American commerce in the early 2020s. Amazon has roughly two million active third-party sellers on its marketplace. The majority of them aren't following Albany's legislative calendar. They're following Amazon's policy emails. And that email arrived Wednesday. The operational shock isn't only about new uploads. Sellers who've already published listings with AI-generated models now face a retroactive question: do existing images need to be re-tagged and re-uploaded? The brief Amazon announcement doesn't appear to draw a hard line on this, and that ambiguity alone will send compliance teams scrambling. For larger catalog sellers running hundreds of SKUs, an audit of every listing for synthetic human content is a real project, not an afternoon task. ## What this means for the AI product photography industry The AI product photography sector has grown quickly on the back of Amazon's own marketplace. Tools from companies including Claid, Pebblely, and a range of newer entrants let sellers swap in AI-generated models for a fraction of the cost of a traditional shoot. For small sellers and overseas manufacturers without access to local photography studios, these tools removed a genuine barrier. The New York law, enforced through Amazon's platform, doesn't ban any of that. It just requires a label. But labels change behavior. Research consistently shows that disclosed AI content is perceived differently by consumers, and sellers who've built listings around the clean, polished look of AI photography now have to decide whether the disclosure indicator Amazon will attach to their listings affects conversion rates enough to justify switching back to real models. Some will calculate that the label is neutral. Others won't want to find out. The broader regulatory trajectory is clear. New York is first, but it won't be alone for long. The Honigman law firm noted in a June 2026 analysis that similar synthetic performer disclosure proposals are moving through multiple state legislatures, and the FTC updated its Endorsement Guides in March 2026 to clarify that undisclosed AI imagery in advertising can constitute a deceptive practice under Section 5 of the FTC Act. Amazon's metadata tagging system, built on C2PA Content Credentials, gives it infrastructure to scale this compliance framework across future requirements without rebuilding from scratch each time. For sellers, the practical answer right now is to audit your catalog for any AI-generated human figures, tag what requires tagging before your next upload cycle, and watch Amazon's seller communications closely over the next few weeks for any clarification on retroactive obligations. The fines are real. The platform is enforcing. And unlike most regulatory surprises, this one came with a policy email and a specific date. **Also read:** [China's memory chipmakers are racing to break the Korean stranglehold on AI infrastructure](https://startupfortune.com/chinas-memory-chipmakers-are-racing-to-break-the-korean-stranglehold-on-ai-infrastructure/) • [China's AI cobots are no longer science fiction and millions of factory workers are about to find out](https://startupfortune.com/chinas-ai-cobots-are-no-longer-science-fiction-and-millions-of-factory-workers-are-about-to-find-out/) • [Singapore tightens cybersecurity rules for critical infrastructure after China-linked hack hit all four major telcos](https://startupfortune.com/singapore-tightens-cybersecurity-rules-for-critical-infrastructure-after-china-linked-hack-hit-all-four-major-telcos/)