Amazon AI video tagging requires a tested container workflow
Amazon's policy language includes applicable video media, but MP4 and MOV store metadata differently from JPG and PNG. AssetDisclose does not yet offer production video writing because the output must be tested across container, codec, and Amazon ingestion combinations.
Why video is not just a larger image
QuickTime-family containers can store XMP and native metadata in several locations. A command that writes a readable tag locally may still fail after another editor, transcode, or platform ingestion step.
What production support must prove
- The tag survives without re-encoding audio or video streams.
- The written value is readable through an independent verification path.
- Common MP4 and MOV container and codec combinations remain playable.
- The metadata remains present after representative marketplace upload tests.
What sellers can do now
Confirm whether the person falls inside the policy, preserve the original video, document the generation source, and use a tested desktop metadata workflow only if your team can verify the final container. Do not assume the current image tool writes video.
Evidence behind this guide
Policy guidance can change. Confirm the current Seller Central instruction before a production upload.
Common questions
Can I upload MP4 or MOV to the current AssetDisclose tool?
No. The production writer currently accepts JPG, JPEG, and PNG only.
Will video tagging change image or sound quality?
A metadata-only implementation should not re-encode media streams, but each supported container combination still needs verification.
How can I join the video beta?
Use the video beta contact link on the AssetDisclose homepage and describe the containers and codecs in your workflow.
Amazon AI media tagging resources
Move from policy interpretation to writing, verification, batch processing, and provenance.