AI avatar for kids videos: made for kids and COPPA, what changes
Marking a video made for kids turns off comments, cards and personalized ads on YouTube. What COPPA covers, and how a Sume avatar clip fits a children audience.
Yes, you can use an AI avatar as the presenter of a kids video, and the avatar changes very little about your obligations. What matters is the audience. If the video is directed to children, you tell YouTube it is made for kids, and YouTube turns off features that collect data. Sume renders the clip; it does not decide your audience setting, and it does not make the content child-safe for you.
The YouTube side comes from its made-for-kids help page and the FTC side from its COPPA FAQ, both read 2026-10-06. This is a summary of those pages, not legal advice.
What changes when a video is made for kids
| Feature | On made-for-kids content |
|---|---|
| Autoplay on home | Disabled |
| Cards or end screens | Disabled |
| Comments | Disabled |
| Personalized advertising | Disabled |
| Live chat or live chat donations | Disabled |
| Channel memberships | Disabled |
| Notification bell | Disabled |
Who decides, and what COPPA looks at
YouTube says creators are responsible for setting the audience, that it uses machine learning to find videos clearly directed to young audiences, and that it can override the choice in cases of error or abuse. It grounds the rules in COPPA, which applies to children under 13.
The FTC says it judges a service as child-directed by the totality of circumstances, including subject matter, visual content, animated characters, child-oriented activities and language. It also says a video publisher on a general-audience platform that collects no personal information and enables no collection on its behalf is not covered by COPPA itself. A bright avatar in a cartoon scene can push the call toward child-directed, so decide on purpose.
What to render with Sume
- Create an adult presenter avatar from a prompt or profile traits, then reuse its handle. Creation is $0.95 once.
- Keep each video to one idea, 4-60 seconds, in 9:16, 1:1 or 16:9.
- Turn on inline captions so the script is readable without sound.
- Write the script yourself and have a parent or teacher read it first.
- Add the AI disclosure where the platform asks for it. YouTube's synthetic-content page requires a label for realistic content that could be mistaken for real, and skips it for clearly unrealistic content such as a unicorn ride.
The short version
Decide the audience first, set it honestly, and then pick the tool. An avatar is a production choice, not a compliance shield.
Sources
Related posts
More in Use cases
- AI avatar objection handling clips for sales reps: 3 examples
Three short answer clips a rep can send after a buyer raises price, security or timing. Scripts sized by 2.8 words a second, the cost, and the review rule.
- AI customer service with a video avatar: where a clip fits
Tavus compared eight AI customer service platforms on 2026-10-02. A live agent answers the long tail; a rendered avatar clip covers top answers. Costs.
- AI music video generator API: which models take your audio?
Seedance, Wan and MiniMax accept audio references on Sume. Gemini Omni and Kling 3 do not. Limits per model for a music-video workflow.
- AI product demo avatar: live agent or recorded clip?
Tavus builds live demo agents that answer buyers; Sume renders a 4-60 second avatar clip with your product image. Where each fits, plus Sume per-second prices.
Written by Sume