YouTube AI label can't be removed: Veo and C2PA cases

YouTube says creators can update a wrongly applied AI disclosure in YouTube Studio, except in a handful of cases where it stays permanent. The blog's list includes two: content made with YouTube's own AI tools, such as Veo, and content containing C2PA metadata that says it is fully generative AI.

When is the label permanent?

What does a Sume file carry?

The Sume docs we read do not say a file carries C2PA metadata, and they do not say it does not. Do not assume either way. Check the file you actually upload. Sume media URLs are durable media.sume.com HTTPS URLs that do not expire and are public to anyone holding the URL; see Runs.

Does trimming or filtering change the answer?

Video trim returns a new MP4 artifact holding only the selected range, and Video filter returns a new MP4 with the source untouched. The docs do not say what either does to embedded metadata, so treat a derived file the same as the original and verify it. This post does not suggest stripping metadata to avoid a label; YouTube still asks creators to disclose realistic AI. See Video trim.

What should I do instead?

Disclose accurately at upload, and use Studio if the label is wrong in a non-permanent case. A label alone does not change recommendation or monetization eligibility, according to the blog. Read what needs disclosure for the rules.