AI footage of a real place on YouTube: the travel promo rule
YouTube's own example: AI-generated extra footage of a real place, such as a surfer in Maui, needs the AI label. How to decide, and the cheap alternative.

Yes: if you generate extra footage of a real place with AI and it looks real, YouTube requires you to disclose it. The GenAI disclosure page gives this exact case: 'AI generated extra footage of a real place, like a video of a surfer in Maui for a promotional travel video'. The same page lists the four things creators must disclose, and a realistic generated scene that did not happen is one of them.
The disclosure does not cost reach or money. The page states that disclosing AI content won't limit a video's audience or impact its eligibility to earn money. The risk sits on the other side: creators who consistently do not disclose can get a label applied manually, or face removal of content or suspension from the YouTube Partner Program.
The four disclosure triggers
YouTube asks you to disclose AI use that meaningfully alters or generates photorealistic content. Read the list against a travel promo before you publish.
| Trigger | Travel promo example |
|---|---|
| Makes a real person appear to say or do something they did not | A named host appearing to endorse a hotel they never visited |
| Alters footage of a real event or place | Adding a generated festival crowd to a real beach |
| Generates a realistic scene that did not occur | A generated surfer on a real Maui break |
| Creates music that is the main focus | A generated song as the whole point of the clip |
Make the compliant version with Sume
You have two honest routes. Route one: generate the shot with Sume video generation, keep it, and answer Yes to 'AI use' when you upload. Pick aspect_ratio and duration from the model's supported_aspect_ratios and supported_durations, which GET /v1/videos/models returns. One model's listing in the docs shows durations of 4 to 15 seconds and 9:16 among its ratios, but read the live list for the model you choose.
Route two: shoot a real clip and use Sume only for non-generative work such as a cut, captions or a crop. The same help page names caption creation, sharpening, upscaling and color adjustment as edits that do not need disclosure, so a real clip with those changes stays outside the rule.
Steps
- Write down which shots are generated and which are filmed.
- If any generated shot looks like a real place or event, answer Yes under Attributes, 'AI use', in YouTube Studio.
- If the generated shot is stylized and obviously not real, the non-realistic exemption may apply; see the page's examples.
- Keep the generation job id with the upload so you can show what was generated.
Caveat
The page says its list is not exhaustive. When a shot is borderline, YouTube's own wording leans toward telling viewers: the requirement exists because viewers want to know if what they watch is real. Disclosing costs nothing in reach.
Sources
Related posts
More in Use cases
- YouTube AI label: player or description, by how real it looks
YouTube shows the altered or synthetic label in the player for photorealistic content, in the expanded description for animated. What it means.
- YouTube AI Shorts tools missing in EEA, UK, Switzerland: use an API
YouTube lists several Shorts AI tools as not available in the EEA, UK or Switzerland. If you are there, generate the clip with an API and upload the MP4.
- YouTube air-gapping: why a filmed screen can't earn the camera label
YouTube's help page names air-gapping: filming a screen that shows synthetic content. Why it can't earn the camera label, and what to do with AI clips.
- YouTube Audio-to-Short: 8 s to 3 min, English transcript required
YouTube's Audio-to-Short turns audio-first videos into Shorts of 8 seconds to 3 minutes. It needs an English transcript and desktop. Prepare both with Sume.
Written by Sume