YouTube AI disclosure exemptions: captions, color and audio repair
YouTube exempts captions, color adjustment, beauty filters, audio repair and own-voice dubs from AI disclosure. Which Sume jobs are only edits.

The exempt list, from YouTube
YouTube's GenAI disclosure page (read 2026-10-09) names edits that do not require a disclosure: primarily aesthetic changes such as beauty filters, color adjustment, voice cloning for one's own dubs, captions and audio repair. It also says production assistance such as script generation, thumbnails, outlines and infographics needs none. The line it draws is whether an edit alters the content in a way that could mislead viewers about what actually happened.
Mapping Sume's media jobs onto that list
Sume's non-generative media jobs run ffmpeg or a caption renderer on a clip you already have. They change timing, framing, brightness or text overlay and do not synthesize new footage. The table maps each documented job to the YouTube category it resembles; whether a specific edit is minor is still your call and YouTube's.
| Sume job | What it does | Closest YouTube category | Public price |
|---|---|---|---|
| Video trim | Cuts [start, end) from one clip | Editing, not generation | $0.02 per job |
| Video filter (dim, crop) | Brightness or crop on one clip | Color adjustment | $0.02 per job |
| Video captions | Burns captions from speech or authored cues | Captions | $0.20 per job, clips up to 60 s |
| Audio detach | Extracts the audio track to wav or mp3 | Audio handling | $0.01 per job |
Where the exemption stops
A caption job does not make a clip exempt if the clip itself is generated. If you made a photoreal scene with a video model and then trimmed and captioned it, the trim and the captions are exempt edits, but the scene is still a generated realistic scene and needs the toggle. Likewise a crop that removes a person from a real event is a different question than a crop to 9:16.
The video-captions docs add a precision worth knowing: speech captions come from speech recognition plus an optional script_text alignment, and authored cues burn your own text with no recognition. In both cases you review the words before publishing; the exemption does not cover wording that misstates what was said.
Own-voice dubs
The page exempts voice cloning for one's own dubs. Sume's docs list text-to-speech priced at $0.0475 per 1,000 characters (catalog read 2026-10-08), with a maximum request of 20,000 characters, $0.95. Do not read that as a statement about cloning: the exemption is about your own voice, and anything that makes someone else's voice say new words is in the disclosure list above.
A short checklist: trims, crops, brightness, captions and audio extraction are edits. A new scene, a swapped face or a synthetic voice of another person is not.
A quick decision rule
Ask one question per clip: could a viewer come away believing something happened that did not? If the answer is no and the only changes are cut points, brightness, framing or burned-in words, the YouTube page treats those as minor. If the answer is yes, disclose, whatever tool made the change.
For batch work, keep two folders: raw generated clips, which carry the disclosure decision, and edited deliverables. A Sume trim returns a new MP4 and never overwrites the source, so the originals stay available when you need to re-check the decision.
Sources
Related posts
More in Use cases
- YouTube's in-Shorts AI clips are 10 s and auto-labelled; yours are not
YouTube's Shorts AI tools make clips up to 10 seconds, auto-disclosed. A clip made on Sume and uploaded is not auto-labelled; you set the answer yourself.
- YouTube's Shorts AI tools self-label; a clip you upload does not
Clips from YouTube's own Shorts AI tools are disclosed automatically and run up to 10 seconds. A clip generated elsewhere and uploaded needs your own toggle.
- Skippable in-stream: no max length, under 3 minutes recommended
Google Ads sets no maximum for skippable in-stream video but recommends under 3 minutes. A 5-minute master becomes a 179-second cut for $0.02 on Sume.
- 3D-Printing Shop Product Clips: Reference Images With Wan 3.0
Show a printed part from several angles without a studio. Wan 3.0 on Sume takes 10 reference images; 4 seconds at 720p costs 50 cents.
Written by Sume