EU AI icon for fully AI-composed music: generated tracks
The EU icons page lists fully AI-composed music or art under the Fully AI-Generated icon, with limited disclosure for artistic works. Handling a Sume track.
Yes, the Commission's icons page names music: its "Fully AI-Generated" row covers content with no human-created elements or editorial control apart from prompting, and the examples include "Fully AI-composed music or art", marked with an asterisk. The asterisk says such works "may enjoy limited disclosure requirements for artistic, creative, fictional or satirical works". Whether your track needs a label at all is a legal question; this is not legal advice.
Icon facts are from the Commission's page, last updated 24 September 2026, read 2026-10-01. Music facts are from Music Router.
What does the page say about when the icon applies?
The icons support Article 50(4), which covers deepfakes and certain AI-generated text on matters of public interest. The table gives three icon types: Basic, Fully AI-Generated, and Partially AI-Modified. The music example appears under Fully AI-Generated. The page says use of the icons is optional but the Article 50 labelling requirements are not, and that using an icon does not establish compliance on its own.
What does a Sume music job give me?
A prompt-driven track. The Music Router public id is sume/music-router, and the routable ids are sume/music-auto (the default), lyria-3.5 and lyria-3-pro. The docs example asks for "A 30-second track. Instrumental, no vocals." The request echoes the model you asked for in job.model and names the engine that ran in job.request.routed_model, which is worth keeping with the file.
| Item | Sume Music Router | EU icons page |
|---|---|---|
| What it covers | Generated audio from a prompt | Fully AI-Generated: no human-created elements apart from prompting |
| Record to keep | job.model and job.request.routed_model | Not requested |
| Artistic works | Not addressed | Asterisk: may enjoy limited disclosure requirements |
Where would the icon go for an audio-only track?
The page's summary says the icon should be directly embedded into the deepfake or published text (except for creative works), unless equivalent alternatives such as a user interface overlay exist, and visible when reshared or downloaded. It does not explain audio-only placement in that summary, so check section 2 of the Code of Practice it links. If the track soundtracks a video, Timeline compose can overlay a still, such as the icon PNG, onto a Sume-hosted video; see the label on an avatar video for the text-label variant.
What should I keep with a generated track?
Keep the job id, job.model, job.request.routed_model and the date next to the file, so you can say later which engine produced it. Prompts are also worth keeping: the Commission's Fully AI-Generated row turns on there being no human-created elements apart from prompting. The asterisk on the music example is about whether disclosure is limited; see artistic works and disclosure. Sume does not decide that for you.
Sources
Related posts
More in Use cases
- Facebook catalog: 20 extra images per product from one photo
Meta lets a catalog item carry up to 20 additional_image_link entries. Make extra angles from your real photo with input_references and n.
- Facebook catalog product video: 20 per item, 200 MB, direct link
Meta catalog items take up to 20 videos of 200 MB each from a direct download link. Make a clip with Sume and hand Meta the Sume media URL, not a player page.
- Faceless education video episodes, 1-5 minutes, on Sume Timeline
Build a 1-5 minute faceless education episode on Sume: one narration track plus scene stills on a Timeline 1.0 document you reuse for each new episode.
- Final Cut Pro convert closed captions: burn cues in Sume
Final Cut Pro 12.3 converts closed captions to subtitles. To burn existing text onto a video elsewhere, send cues or segments to Sume; SRT is unsupported.
Written by Sume