YouTube's AI Shorts tools are missing in the EU: what to use

Most of YouTube's AI features for Shorts are unavailable in the European Economic Area, and photo-to-video and Make Me Move are also unavailable in the UK and Switzerland. Where they do work, YouTube says the tool discloses AI use for you automatically. If you are outside the supported regions, you can generate a clip elsewhere, but then you have to set the disclosure yourself.

This post reads two YouTube help pages (AI-generated features for Shorts, and disclosing altered or synthetic content), both read 2026-10-02.

Which AI tools does YouTube offer for Shorts?

The AI-features page lists five, with these availability notes.

Does a language setting matter?

Yes. The page says the features generally require the YouTube device language to be set to English. It also says YouTube's safeguards block prompts that violate policy or touch sensitive topics, and that the system avoids generating photorealistic images of identifiable people.

Do I have to label a Short made with YouTube's own AI tools?

No extra step. The page states that creators who make a post or Short using YouTube's generative AI tools do not need to take extra steps to disclose the use of altered or synthetic content, because the tool automatically discloses it.

The label rules for everything else are on the disclosure page. You disclose realistic content that makes a real person appear to say or do something they did not, alters footage of a real event or place, or generates a realistic scene that did not occur. You choose Yes under "AI use" in the Attributes section of Studio's upload flow. Non-realistic content and minor edits do not need it.

What if I make the clip outside YouTube, for example with Sume?

Then the automatic disclosure does not happen through YouTube's tool, so the decision is yours at upload. The disclosure page also says YouTube may label content that contains C2PA metadata or that its systems detect as AI-generated, so a label can appear even when you chose No. For how permanent that is on YouTube's own tools, see the Veo and Dream Screen post.

Sume's video generation is an API, so the clip is generated by a request rather than a Studio screen. Auto create controls default to 720p and 8 seconds, with 3-10 second clips at 16:9 or 9:16. I found nothing in the docs that ties a request to a YouTube country list, but check the model and your own legal setting yourself. A 9:16 request:

curl -X POST https://api.sume.com/v1/videos \
  -H "Authorization: Bearer $SUME_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: vertical-clip-001" \
  -d '{
    "model": "sume/auto",
    "prompt": "A slow push-in on a ceramic mug on a wooden desk, morning light",
    "aspect_ratio": "9:16",
    "duration": 8
  }'

What is the safe workflow outside the supported regions?

Generate the clip, trim or caption it as needed, answer the AI-use question honestly on upload, and keep the file. If the clip shows a realistic person or event that did not happen, answer Yes. Sume does not disclose for you on YouTube. For the full disclosure boundary, read what needs disclosure on YouTube.

What should I record for my own files?

Keep a note of which tool generated each clip, the prompt, and the date. If a clip is later questioned, that note is how you show what was real and what was generated. YouTube's page says its own tools disclose automatically, but a clip made elsewhere carries only the answer you gave.