Amazon Now Requires Disclosure of AI-Generated People in Product Images — What It Signals

What the Rule Actually Requires

Sellers must add a designated disclosure keyword to the file of any product image or A+ Content that features a photorealistic AI-generated person. The scope is specific: it targets synthetic humans presented as if real, rather than AI-assisted background removal, color correction, or other editing applied to genuine photography. The distinction matters, because a great deal of routine e-commerce image work now involves AI tools without generating synthetic people.

Why Amazon Moved on This

Photorealistic AI models have made it trivial to generate convincing images of people wearing, using, or endorsing products who don't exist. For a marketplace built on shopper trust, undisclosed synthetic models create a credibility problem — a customer who believes they're seeing a real person's fit or use of a product is being shown something fabricated. Requiring disclosure is a way to preserve trust without banning a technology that's already widespread.

The Broader Signal for E-Commerce

Amazon's scale means its content standards tend to become de facto industry norms. Expect other marketplaces, ad platforms, and potentially regulators to adopt comparable transparency requirements for synthetic imagery. Brands that treat this as an Amazon-only compliance chore risk being caught flat-footed when the same expectation arrives on every channel they sell through.

• — Tag synthetic-person imagery consistently across all channels, not just Amazon
• — Keep an internal record of which assets are AI-generated versus photographed
• — Distinguish clearly between AI-edited real photos and fully synthetic people
• — Treat disclosure as a trust feature, not merely a rule to satisfy

Does This Mean AI Imagery Is a Bad Idea?

No — it means undisclosed synthetic humans are becoming a liability while transparently labeled AI imagery remains a legitimate tool. Plenty of product photography benefits from AI assistance in ways that involve no synthetic people at all. The emerging line is between using AI to enhance real assets and using it to fabricate people who appear to be real customers or models.

Building Disclosure Into the Workflow

The practical move is to add provenance tracking to your image pipeline now: knowing, for every asset, whether it's a photograph, an AI-edited photograph, or a fully generated image. That record makes any platform's disclosure requirement a simple export step rather than a forensic audit of thousands of existing images later.

Managing Product Assets at Scale — With Devrex Digital

Devrex Digital builds custom stores with structured product asset management — including provenance metadata that tracks how each image was created — so disclosure requirements from any channel become a straightforward export rather than a crisis. If your catalog mixes photographed and AI-generated imagery without a clear record of which is which, that's worth fixing before more platforms adopt rules like Amazon's.