DRULES AI
🏠 Home 📰 Blog
← All posts

AFM Sues UMG & WMG Over Suno-Udio AI Payouts

The American Federation of Musicians has sued Universal Music Group and Warner Music Group, alleging the labels licensed member recordings to Suno and Udio without paying or crediting the musicians who performed on them.

🎯 Lawsuit Targets Label Greed

Filed in early June in New York federal court, the AFM complaint claims UMG and WMG received substantial compensation through retrospective settlements and prospective licensing deals with the AI firms yet shared none of it with the 70,000 union members whose work sits in the training datasets. The suit accuses the labels of breaching collective bargaining agreements that require compensation and credit when recordings are licensed.

Warner settled its Suno litigation in November 2025, licensing its catalog for future AI training with artist opt-in promises. UMG maintains an active suit against Suno while reportedly striking similar commercial arrangements. The AFM argues these deals monetize musicians' past performances without flowing revenue to the actual creators.

📜 Settlement Backdrop Exposed

The legal action arrives amid fresh scrutiny of exactly what data Suno and Udio consumed. Recent reporting highlights searchable databases containing roughly 21 million tracks used for training, including chart hits. Sony, UMG and Warner originally sought statutory damages up to $150,000 per infringed work.

While labels positioned themselves as victims in 2024 suits against the AI companies, the AFM lawsuit flips the script: the majors protected their own interests and created new revenue streams but left session musicians out. Posts from composers and producers on X amplified the outrage, with one prominent creator reminding Suno of its earlier court admission that it trained on "essentially all music files of reasonable quality" scraped from the internet.

🔊 Community and Industry Fallout

Reactions split along predictable lines. Independent artists see validation for long-held concerns about unauthorized training. Some point to fair-use precedents like the Bartz v. Anthropic ruling that purchasing tracks for transformative use may be legal, but the scale and lack of consent still anger working musicians.

Meanwhile, Suno users continue experimenting with the platform's newly available v4.5-all free model, generating tracks while the corporate battles rage. The AFM suit demands accountability beyond label coffers, seeking to force revenue sharing and proper crediting for any future AI licensing.

Industry watchers note this could set precedent for how unions extract value from generative tech. Parallel efforts include tools like HarmonyCloak that add inaudible perturbations to recordings, making them resistant to AI training.

Bottom line: Session musicians are no longer content watching labels cash in twice on their performances—once on the original release and again on AI training licenses.