Does a scripted AI presenter video need YouTube's synthetic label?
YouTube asks for a label when content makes a real person appear to say something they did not. Here is how to read that for a made-up Sume Avatar presenter.

Short answer
It depends on whom the presenter looks like and what the video shows. YouTube's help page asks creators to disclose realistic altered or synthetic content, and its first example is content that makes a real person appear to say or do something they did not do. A presenter invented from a text prompt is not a real person, but a photorealistic one can still read as a generated scene that did not occur. We cannot give you a legal or policy ruling; this post sets out the page's own categories and how to test your video against them.
The four triggers on the YouTube page
The page lists what to disclose when content is realistic. Compare each against your video before you upload.
| Trigger | Applies to a made-up presenter? |
|---|---|
| Makes a real person appear to say or do something they did not | No, if the face is not a real person; yes if built from a real person's photo without them saying it |
| Alters footage of a real event or place | Only if your scene photo is a real place you changed |
| Generates a realistic scene that did not occur | Possibly: judge whether a viewer would take it as real footage |
| Music that is the main focus | Not for a spoken presenter |
What the page says you can skip
The page lists production help as exempt: script generation, thumbnail creation, caption generation, and video enhancement such as upscaling. It also lists fantasy content and cosmetic edits. In a Sume workflow, writing the script with a model and adding captions with the captions feature fall in those exempt groups. The avatar render itself is the part to assess.
- Exempt on the page: scripts, thumbnails, captions, sharpening, upscaling.
- Fantasy or clearly unrealistic content needs no label.
- The avatar render is the step to judge.
How Sume inputs map to the question
Sume creates avatars from a prompt, structured profile traits, or a reference image. A prompt or profile avatar has no real-person source. An image avatar built from a photo of a real person is different: if that person did not say the script, the first YouTube trigger is in play. For the scene, a prompt scene is generated, while a photo scene uses your own image.
| Input | Sume field | Real-person question |
|---|---|---|
| Prompt avatar | input.type prompt | No real person |
| Profile avatar | input.type props | No real person |
| Image avatar | input.type photo | Yes, check consent and the script |
| Photo scene | scene.type photo | Real place: do not alter it deceptively |
What happens if you choose no
The page warns that creators who consistently choose not to disclose may get a manual label or penalties, including content removal or suspension from the YouTube Partner Program. It also says that disclosing does not limit audience reach or affect monetization eligibility. When a case is unclear, the page's own wording makes disclosure the lower-risk choice.
A practical test before you upload
Ask three questions of the finished file. Would a viewer reasonably take the presenter for a real, identifiable person? Does the script put words in that person's mouth that they never said? Does the background pass as real footage of a real place? A yes to any of them is a reason to disclose. If all three are no, for example a clearly stylized presenter introducing your own product, the page's fantasy and unrealistic-content exemption may apply, though the safest habit remains to label.
Keep a short record next to the file: the avatar input type, the script text, who approved the first frame, and the label decision. If you ever have to explain a video, that record answers the question in a minute.
Sources
Related posts
More in Sume Avatar 1.0
- Will a YouTube AI label hurt reach? What the page says
YouTube's help page says disclosing AI content does not limit reach or monetization eligibility; penalties target non-disclosure. Plan an avatar series on that.
- Introducing Sume Avatar 1.0
Sume Avatar 1.0 is a multi-agent orchestration system as a single avatar model.
- Avatar Face Swap API (Beta): apply an avatar face to a video
Avatar Face Swap 1.0 is a Beta Sume endpoint that applies a ready avatar's face to a short public source video. Required fields, limits, and polling.
- Avatar video previews: approve the first frame before rendering
Create an avatar video preview to get first-frame stills, regenerate them if needed, then call generate-video on the preview id to render the final video.
Written by Sume