Amazon announced on July 23, 2026 that sellers must tag any product image or video containing photorealistic AI-generated people before uploading it to listings or A+ Content across all worldwide stores. The requirement is effective immediately, with no grace period announced.

The policy applies globally. If your image or video contains a photorealistic AI-generated person, you must add the keyword “contains-synthetic-performer” to the XMP dc:subject metadata field of the file before uploading it to Seller Central. Amazon will then add a visible indicator to the image on the product page, notifying shoppers that the content includes AI-generated people.

Why This Is Happening Now

Amazon's announcement ties the requirement directly to legislation. Multiple jurisdictions have passed or are enforcing disclosure laws for AI-generated content in commercial contexts, and Amazon's policy positions the platform ahead of those enforcement deadlines.

The most immediate legislative driver is the EU AI Act, which mandates machine-readable metadata on AI-generated content used in commercial contexts, with compliance required by August 2, 2026. That deadline is nine days from the date of Amazon's announcement. Separately, the FTC established a dedicated AI enforcement unit in January 2026 and increased maximum penalties for disclosure violations to $53,088 per violation, with each non-compliant piece of content counting as a separate violation. California and New York have also enacted state-level AI disclosure requirements that apply to commercial advertising content.

Amazon is not the first platform to move on this. Meta requires “AI-generated” or “Made with AI” labels for synthetic people in advertising content and uses IPTC metadata detection to apply labels automatically. TikTok requires disclosure of all realistic AI-generated content via a built-in label. Amazon's approach places the tagging responsibility explicitly on the seller rather than relying on automated detection.

What You Need to Do

The requirement applies specifically to photorealistic AI-generated people. You do not need to tag images that feature real people, even if those images have been AI-edited or enhanced. You also do not need to tag images that show characters from movies, TV, or video games, images without any people, or images with non-photorealistic AI-generated figures.

The tagging method varies by operating system. On macOS, open the file in Preview, go to Tools, then Show Inspector, click the Keywords tab, and add “contains-synthetic-performer” as a keyword. On Windows, right-click the file, go to Properties, then the Details tab, and type “contains-synthetic-performer” into the Tags field, not the Subject field. Both fields look similar but write to different metadata locations, and only the Tags field is correct. For Linux or command-line workflows, Amazon's help page provides exiftool commands that append the keyword to the XMP-dc:Subject field without overwriting other existing metadata.

After tagging, verify the metadata is present before uploading. Any XMP-capable reader can confirm the contains-synthetic-performer keyword appears in the dc:subject field.

What This Means for Your Image Workflow

AI-generated lifestyle imagery has become a standard tool for sellers looking to produce product photography at lower cost than traditional studio shoots. Generative AI tools now produce photorealistic human models in seconds, which many sellers use for apparel, beauty, home goods, and other categories where lifestyle imagery featuring people drives conversion.

The new tagging requirement adds a step to that workflow but does not prohibit AI-generated people in listings. You can continue using AI-generated models in your images and A+ Content; you just need to tag the files before upload and accept that Amazon will show a disclosure indicator to shoppers on the product page.

Whether that disclosure indicator affects conversion rates is an open question. Shoppers already see AI-generated people in advertising across every major platform, and the label may carry little practical weight in most categories. In categories where authenticity and trust are particularly important, such as skincare, fitness, or any category where consumers are assessing how a product looks on real people, a visible AI disclosure on a model image is worth thinking through carefully before making it your default.

The broader trend across platforms suggests this is only the beginning of AI content disclosure requirements. Meta, TikTok, Google, and YouTube all implemented AI labeling requirements through 2025 and into 2026, and Amazon's policy brings its product listings into alignment with what those platforms already require for advertising. Sellers building AI-generated imagery into their standard listing workflow should add metadata tagging as a non-negotiable step in that process, both to comply with Amazon's policy and to stay ahead of disclosure requirements that are tightening across every commercial content channel.