Topic

Training-data

Suno admits it kept the training recordings, and v6 learned from its users' creations
In its answer to Universal and Sony's second lawsuit, Suno admits it has not deleted all the recordings it trained on, and repeats that v6 learned from "creations and preference signals" from its community. That makes Suno's own users part of the training pipeline the labels call laundering, under a licence they can't take back.
23 Oct 2026Artist rightsvia US District Court for the District of Massachusetts (via Music Business Worldwide)
The Copyright Office's AI training report is still a draft 17 months on, and its critics are using that
The US Copyright Office's report on generative AI training, released as a "pre-publication version" on 9 May 2025, has never been finalised. Its view that some training is fair use and some isn't, and that market dilution is real harm, is the closest thing music has to an official US position. Litigants cite it, the Justice Department attacks it, and the two big 2025 rulings didn't cite it at all.
16 Oct 2026Artist rightsvia US Copyright Office
The DOJ says AI training is fair use. Music's best arguments were never about training
The US Justice Department has told the court in the New York Times' case against OpenAI that training AI models on copyrighted text is fair use, and it goes out of its way to reject the "market dilution" theory the music industry leans on. It leaves outputs, data sourcing and identity open, and that is where artists' claims are strongest.
2 Sept 2026Artist rightsvia US Department of Justice (copy hosted by MusicTechPolicy)

Suno admits it kept the training recordings, and v6 learned from its users' creations

The Copyright Office's AI training report is still a draft 17 months on, and its critics are using that

The DOJ says AI training is fair use. Music's best arguments were never about training

Subscribe to We are the music makers