Sony and UMG sue Suno again, alleging v6 'launders' stolen music
Sony and Universal Music Group have filed another lawsuit against Suno, the AI music generation company, over its latest v6 model. The labels claim v6 still infringes their copyrights even though Suno says it was built with a new training process, because that process relied on user outputs from Suno's earlier models, which were trained on unlicensed music pulled from YouTube and other sources. Sony and UMG are described as notable holdouts that never signed a licensing agreement with Suno, unlike some other rights holders.
In the complaint, shared with The Verge, the labels accuse Suno of what they call 'model laundering,' arguing: 'Training a new model on the outputs of an infringing model does not eliminate the infringement; it launders it, passing the value of Plaintiffs' expression from the copied recordings into the tainted models, from those models into their outputs, and from those outputs into v6 ... V6 is not a fresh start; it is the fruit of the same poisoned tree.'
When v6 launched, Suno's Jack Brody told The Verge the model was 'trained from the ground up, with a new set of data,' including 'user data,' without giving further detail. Suno spokesperson Rachel Racusen later told The Verge that this data includes 'creations' from users, saying v6 'was trained on content licensed from our partners, interactions including creations and preference signals from our community, and the accumulated learnings from our team.' Racusen did not specify whether that included uploaded audio itself or only outputs generated from uploaded audio.
Sony further alleges that Suno used distillation to train v6, meaning the new model was trained to replicate the outputs of its previous 'teacher' models, which were themselves created using infringing data. The complaint states that 'even a model not directly trained on Plaintiffs' recordings is informed by, and benefits from, Suno's retained unauthorized copies.' The labels' core argument is that unless Suno truly starts from scratch, its models will keep being built on stolen data.
Key facts
- Sony and Universal Music Group filed a new lawsuit against Suno over its v6 model, following prior litigation.
- The labels accuse Suno of 'model laundering': training v6 on outputs from earlier Suno models that were trained on unlicensed music.
- Sony alleges Suno used distillation, training v6 to replicate results from earlier 'teacher' models built on infringing data.
- Suno has said v6 was 'trained from the ground up' using licensed content, user creations, and other data, but has not detailed whether uploaded audio was included.
- Sony and UMG are described as holdouts that never signed a licensing agreement with Suno.
Why it matters
The lawsuit tests a new legal theory in AI copyright disputes: that a company cannot clean up an infringing model simply by training a 'new' one on that older model's outputs. If courts accept the 'model laundering' argument, it could make it much harder for AI companies to claim a fresh start by retraining on their own systems' generated content rather than the original disputed material.
Who it affects
Suno faces continued legal exposure from two major labels that never licensed their catalogs to it, unlike other rights holders who have settled or signed deals. The case also matters to musicians and rights holders watching how courts treat AI systems trained indirectly on copyrighted work, and to other AI music and content generators using similar retraining or distillation approaches.
How to use it
There is no product action here for readers; this is ongoing litigation, not a feature or release with terms to adopt.
How solid is it
The account is based on the actual complaint, shared directly with The Verge, plus on-record statements from Suno's Jack Brody and spokesperson Rachel Racusen. The source does not name the court, filing date of the suit itself, or any damages sought, and does not specify which earlier Suno model versions generated the training outputs in question.
Risks and caveats
Suno maintains v6 was trained on licensed content and user data 'from the ground up,' but has not clarified whether that data included uploaded audio or only outputs derived from it, leaving the central factual dispute over what exactly went into v6's training unresolved.
“Training a 'new' model on the outputs of an infringing model does not eliminate the infringement; it launders it, passing the value of Plaintiffs' expression from the copied recordings into the tainted models, from those models into their outputs, and from those outputs into v6 ... V6 is not a fresh start; it is the fruit of the same poisoned tree.”
— Sony and UMG, from their complaint