Real footage plus one AI shot: do you tick YouTube's AI use label?

Probably yes, if the AI shot is realistic and fits one of YouTube's four tests; no, if it is clearly unrealistic or only a minor aesthetic edit. YouTube's disclosure page asks creators to disclose realistic altered or synthetic content that makes a real person appear to say or do something they did not, alters footage of a real event or place, generates a realistic scene that did not occur, or creates music that is the main focus of the video (read 2026-10-03).

The page describes one "AI use" setting in YouTube Studio and gives no per-shot option or rule for mixed cuts. So the practical test is the cut as a whole: does any realistic shot in it meet a test? If one does, disclose the video.

How do I sort each shot in the cut?

Go through the timeline shot by shot and mark each against the page's lists.

Where does the label show, and who else can add it?

Per the page, the label appears in the video player for photorealistic AI content and in the expanded description for non-photorealistic or animated material. YouTube also applies labels itself for content made with its own generative AI tools, content with C2PA metadata, and content its systems detect as AI-generated. So a label can appear even if you did not tick the setting.

How does a mixed cut come out of Sume?

Timeline 1.0 takes Sume-hosted clips in ordered video[] slots and returns one MP4 (docs read 2026-10-03). A clip you shot yourself has to be imported first; video inspect notes there is no open-internet fetch and points to POST /v1/media-imports. The render does not carry a per-shot record of which clip was generated, so keep your own shot list.

The disclosure itself is not part of the file. You set it when you upload, in Studio or in the API field covered by the last related post. Marking the AI shots in your shot list before you export is the step that makes the Studio decision a two-minute one.