Sony and Universal Music Group sue Suno again over alleged copyright infringement in v6 model

Sony and Universal Music Group (UMG) filed a lawsuit against Suno, alleging that the company’s v6 model infringes their copyrights through a practice the labels describe as 'model laundering.' In the complaint, the labels argue that v6 was trained on outputs from earlier Suno models, which were themselves trained on unlicensed music sourced from platforms like YouTube. Sony and UMG, who did not sign a licensing agreement with Suno, claim this chain of training perpetuates copyright infringement by passing the value of their copyrighted recordings into v6 and its outputs.

The complaint specifically alleges that Suno used a technique called distillation to train v6, where the new model replicates the results of prior 'teacher' models created using infringing data. Even if v6 was not directly trained on the labels’ recordings, Sony and UMG argue it benefits from Suno’s retained unauthorized copies of copyrighted works. The labels assert that unless Suno’s models are built from scratch with fully licensed data, they will continue to be 'the fruit of the same poisoned tree.'

Suno responded to the allegations in a statement to The Verge, asserting that v6 was trained on a 'new set of data' that included 'user data.' A spokesperson later clarified that the training data comprised 'content licensed from our partners, interactions including creations and preference signals from our community, and the accumulated learnings from our team.' However, the spokesperson did not specify whether this included uploaded audio or outputs derived from such uploads.

The lawsuit follows prior legal action by Sony and UMG against Suno and reflects broader industry efforts to address alleged copyright infringement in AI-generated music. The outcome of this case could influence how AI companies approach model training and data sourcing in the music sector.