Suno released Suno v6, a new family of models, on Wednesday, and the company says none of it was trained on the disputed data behind its earlier models. That distinction, drawn explicitly by Suno itself, is the actual news: a generative music company rebuilding its core product on licensed catalogs while its legal exposure from the old approach is still working through the courts.
The licensing sources are Warner Music Group, BMG, and Believe. Warner reached a settlement with Suno last year; BMG signed a licensing deal last month. Sony Music and Universal Music Group have not, and both are still suing. Musician Jason Isbell has a separate suit pending, and a group of users has filed a class action alleging Suno mishandled a data breach. None of these cases has produced a finding of infringement. Suno has settled with one label and licensed with another, and it disputes the remaining claims.
The timing compounds the contrast: Suno’s v6 announcement landed one day after the company acknowledged it trained earlier models using audio pulled from YouTube. Read together, the sequence looks less like vindication and more like triage, a company drawing a hard line between its legacy training data and everything it ships from here forward.
Three tiers make up the new lineup. The base Suno v6 model targets paying users who want controllable, reproducible output. Suno v6-wild, also paid, is pitched at ideation and open-ended results rather than precision. Suno v6-mini is free and faster, aimed at breadth of usage rather than fidelity. The older models are being retired, which forces the paying base onto the licensed foundation whether or not the pending suits resolve in Suno’s favor.
The product layer is where the licensing decision starts to look less like risk management and more like a business model. Suno is building a remix feature that requires artists to opt in, and it is setting up a revenue-sharing arrangement with labels for AI-generated derivative works. Suno’s chief product officer, Jack Brody, told TechCrunch the goal is new income for rights holders: “I think the music ecosystem and our partners are always looking for ways to create more revenue opportunities for their rights holders and artists.” That is a bet that the labels Suno hasn’t yet settled with, Sony and Universal chief among them, will eventually prefer a cut of AI output to a courtroom.
Brody also addressed platform abuse, telling TechCrunch that Suno can help with “streaming fraud and kind of mass exportation and uploading to distributors that are low intent,” while placing ultimate gatekeeping on distributors and platforms rather than Suno itself. The company has also added a watermark for generated tracks and tightened download limits by account tier, both moves aimed at the same downstream problem: AI-generated music entering the same distribution pipes as human-made work without disclosure.
None of this resolves the outstanding suits from Sony, Universal, and Isbell, and Suno’s YouTube-training admission gives those plaintiffs fresh material rather than less. What has changed is that Suno no longer treats licensing as purely a legal defense. It is now a marketing claim, a retention mechanism for the paid tiers, and the basis of a planned revenue split with labels. Companies elsewhere in generative media, particularly video and voice, are watching whether that framing holds up against courts that haven’t yet ruled: if Suno can convert a licensing deal into a feature customers pay for, expect competitors training on unlicensed catalogs to face pressure to match the model, independent of how the pending litigation resolves.
Reported by Ivan Mehta for TechCrunch, published September 9, 2026.
