Amazon AI-generated people tag: contains-synthetic-performer
Amazon's seller post says photorealistic AI people in listings, A+ and shoppable videos need the XMP keyword contains-synthetic-performer. Tag the final file.
According to a Seller Central forum post, Amazon now requires the keyword contains-synthetic-performer in the dc:subject (XMP) field of images and videos that contain photorealistic AI-generated people. It covers listing images, A+ content images and videos, and shoppable videos. Because the tag lives in the file, add it to the last file you produce, after any trim or Timeline render.
The rules below are as written in that forum post, read 2026-09-30; check Seller Central for the current policy.
What exactly does the tag cover?
The post names an IPTC-compatible metadata editor as the way to set the keyword, and says the scope is all worldwide Amazon stores.
| Needs the tag | Does not need the tag |
|---|---|
| Product listing images with photorealistic AI people | Real people only, even if altered with AI tools |
| A+ content images and videos | Characters from movies, TV or games |
| Shoppable videos | Images without people |
| Non-photorealistic people |
Why tag the final file and not the generated one?
Any step that renders a new file starts a new file. Timeline "returns **one** MP4" built from your clips, and a default trim is a "frame-accurate re-encode" that produces a new MP4 rather than editing the source. Generated media is served from Sume-hosted media.sume.com URLs, so the practical order is: finish editing, download the final MP4, then write the keyword with your metadata tool. See Timeline 1.0 and Video trim.
Sume's video inspect surface only probes and samples stills; it is not a metadata writer, so use it to check duration and format, not to set tags.
Which Sume outputs show AI people?
Avatar videos are described as a "script-driven talking video" from a generated presenter, so a realistic presenter is the case the tag question is about. Generated video and images with realistic people are the same question. Whether a given face is photorealistic is your call against Amazon's wording.
What are sellers reporting?
The thread says tags disappear after Amazon's image processing, that Seller Central has no verification mechanism, and that it is unclear whether the rule applies to existing content. Amazon replied that it was working with the relevant teams. Keep the tagged master file so you can re-upload it. Sume does not claim a tagging or provenance feature on the docs pages cited here; see does Sume add C2PA or a watermark.
Sources
Related posts
More in Use cases
- AI Act Article 50(2): 2 December 2026 grace period, in plain terms
Commission FAQ: providers of systems on the market before 2 August get until 2 December 2026 to mark outputs. Older content needs no retroactive label.
- ASA on AI deepfake ads: you answer for the output you run
CAP says a realistic AI celebrity endorsement can breach the Code and the advertiser stays responsible. Which Sume avatar inputs involve a real person's photo.
- Black Friday countdown video ad: burn the days-left text
Meta says countdown overlays create Black Friday urgency. A countdown is authored text, so burn a new days-left cue onto one clip each day with Sume.
- Bluesky video captions: WebVTT sidecar vs burned-in
Bluesky video embeds take WebVTT caption files next to the video. Sume burns captions into pixels, so build the VTT yourself from sentence segments.
Written by Sume