On September 18, 2026, Sony Music Entertainment and Universal Music Group filed a 45-page complaint with the U.S. District Court for the District of Massachusetts, alleging that the Suno v6 model infringes 60,202 sound recordings.
Reconstructing the Facts
The suit is a continuation of the initial complaint filed in June 2024. The complaint states that although Suno v6 has reached licensing agreements with Warner Music, BMG and Believe, the new model was partly trained on outputs from previously unlicensed models. UMG and Sony Music contend that this amounts to "laundering" historically infringing data and constitutes continued exploitation of copyrighted recordings. The complaint names 60,202 recordings and asks the court to confirm infringement and award statutory damages and attorney's fees.
According to the complaint, Suno acknowledges that training the new model used the outputs of prior models. On that basis, UMG and Sony Music argue that every model generation stems from unauthorized reproduction, and that v6 is not a fresh start.
Breaking Down the Mechanism
The core logic of the complaint rests on the "fruit of the poisonous tree" theory: even if a new model secures licenses from some rights holders, if its training data comes from the outputs of previously infringing models, it still extends the original chain of infringement. UMG and Sony Music note that Suno retained and continued to use unauthorized copies generated by earlier models, leaving every model version directly linked to the infringing recordings.
This mechanism exposes a compliance gap in generative AI training data: licensing only a new version cannot automatically erase the traces of infringement left by earlier training processes. The complaint stresses that Suno's earlier scraping of data from public sources was already challenged in the 2024 lawsuit.
Industry Impact
The lawsuit directly affects the licensing path of the AI music generation industry. Warner Music, BMG and Believe have partnered with Suno to launch v6, while UMG and Sony Music have not, underscoring divisions among record companies over AI collaboration. The complaint notes that Suno's earlier licensing deals were seen as an acknowledgment of market demand for licensing copyrighted works, but they failed to resolve the historical data problem.
Similar lawsuits have expanded to other parties, including cases brought by Round Hill Music and Jason Isbell. UMG has also recently taken action against Anthropic and DistroKid, reflecting its broader regulatory stance on AI-generated content.
Strategic Assessment (Analysis)
Based on the available facts, licensing a new model can hardly sever the historical chain of infringement completely, and may force AI companies to reassess the compliance of their training data sources. If the court adopts the "fruit of the poisonous tree" argument, the industry may need to more thoroughly isolate or retrain early models, which would raise compliance costs and reshape the negotiating landscape with rights holders.
The case highlights the limits of partial licensing agreements in the field of generated content. Stakeholders must weigh the trade-offs between continuing to support existing models and fully rebuilding their data pipelines, while historical precedent shows that problems of this kind in the infringement chain often lead to prolonged litigation and regulatory scrutiny.
© 2026 Winzheng.com 赢政天下 | 转载请注明来源并附原文链接