AI music generators (Suno, Udio): can I release the songs and who owns them?

I have been sketching ideas in Suno and one of them is honestly good. Can I put it on Spotify? Who owns it? And what is the state of the lawsuits, because I do not want my release pulled next year.

Three layers, and they point in different directions.

The tool's terms: on the paid plans, Suno and Udio both let you use the output commercially; the free tiers generally do not. So the first check is simply which plan made the track.

Copyright: under the current US Copyright Office position, a recording generated entirely by AI has no copyright owner. You can release it, but you cannot stop someone else from releasing the same or a similar output, and you cannot register it as yours. The more of your own work is in the final track (your lyrics, your vocals, your arrangement and edits), the stronger your claim over that part.

The industry: the big labels sued both companies in 2024 over training data. Since then several of those disputes have moved toward licensing deals rather than shutdowns, and the tools have kept running throughout. That reduces the "pulled next year" risk, but the terms of what you can do keep shifting with each deal, so re-read them when you release.

Distribution: distributors and Spotify want AI involvement disclosed, and they remove tracks that impersonate real artists. A release with an AI backing track and your real vocals and lyrics is the safest, most honest shape.

Practical advice from the music side of me: use the generated version as the sketch, re-record what you can, and keep the project files. If a dispute ever comes, being able to show your own work is the whole defence.