YouTube likeness detection: setup, limits, AI avatars
YouTube likeness detection finds videos that alter or fake your face. Who can enroll, what it scans, review actions, and what it means for avatars.
YouTube likeness detection is an experimental tool that lets a verified creator find videos where their face appears to be altered or generated by AI, then review each match and choose what to do. It is for people protecting their own face. It does not scan for AI avatars in general, and it currently matches faces only, not voices.
Everything about YouTube below is from YouTube Help, read 2026-09-29. The Sume side is the avatar creation docs.
Who can set up likeness detection?
YouTube calls it experimental and says it is not available in some countries. You must be over 18 and a Channel Owner or Manager, and you must verify with a government-issued ID plus a short selfie video. The selfie video also serves as the reference the system uses to look for your face. YouTube says verification may take up to 5 days.
What does it scan, and what can I do with a match?
The system runs a one-time search of newly uploaded videos for the face of each enrolled creator, in the way Content ID searches for copyrighted content. Setup and review live in YouTube Studio under Content detection, then Likeness. For each match you have three choices.
| Action | When YouTube says to use it |
|---|---|
| Submit a likeness removal request | You think the video violates the privacy policy; it goes through the privacy complaint process |
| Submit a copyright removal request | Your original copyrighted content was used without permission |
| Move to archive | Remove it from your review list and keep it live; you can start a complaint later |
Does it detect a voice clone?
Not yet. YouTube says the feature currently detects matches of an enrolled creator's face and aims to extend to audio in the near future. For a voice you believe violates the privacy policy, it points to the privacy complaint process. Matches can also include real footage: in this experimental phase the tool may show short clips of your own videos uploaded by other channels, which cannot be removed under privacy policies but may qualify for a copyright request.
What does this mean if I make an AI avatar?
Two things. First, never build an avatar on someone else's face: the person can enroll and have it flagged. Sume's avatar creation accepts a text prompt, structured traits (props) or a reference photo, and the docs page checked describes only those inputs and does not mention a consent step, so getting the subject's agreement is yours to do. Second, enrollment is the person's own choice: YouTube stores the reference data and lets you turn the tool off, after which it can take up to 24 hours to stop finding matches.
For the policy side of publishing avatar videos, see YouTube's AI content policy and AI avatars for YouTube videos.
Sources
Related posts
More in Sume Avatar 1.0
- 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.
- How to create a reusable AI avatar with the Sume Avatar 1.0 API
Send POST /v1/avatar-1.0/generate with an avatar_handle and a prompt, profile, or image input. Poll the job, then reuse the handle for avatar videos.
Written by Sume