AFM Amends Suit Over Suno Udio Licenses
The American Federation of Musicians filed an amended complaint on July 24 against Universal Music Group and Warner label entities, alleging that recordings by union members were licensed to Suno and Udio without compensation or credit. The complaint argues that AI-training licenses trigger the Sound Recording Labor Agreement's "new use" provisions and that the labels failed to provide required licensing information.
The American Federation of Musicians (AFM) filed an amended complaint on July 24 in the Southern District of New York against Universal Music Group, Warner Records, and Atlantic Recording Corp. The union alleges that recordings featuring AFM-represented musicians were licensed to AI music companies Suno and Udio without compensation or credit, in breach of the Sound Recording Labor Agreement (SRLA).
According to the amended complaint, the union also alleges that the labels have not provided information identifying the recordings licensed or the musicians whose work was involved. The AFM seeks damages and information related to the Suno and Udio agreements.
The amended filing replaces Warner Music Group Corp. with Warner Records and Atlantic Recording Corp. as defendants. Digital Music News reported that Warner Music Group had argued the parent company was not a signatory to the SRLA, did not own copyrights, and did not itself hold licenses with Suno or Udio.
The contract dispute centers on "new use"
Music Business Worldwide reports that the AFM bases its claim on Article 21 of the SRLA, a collective bargaining agreement covering sound recordings. The union argues that the provision requires signatory labels to compensate musicians and notify the union when recordings are used for purposes not covered by the agreement.
The complaint characterizes the licensing of recordings for generative AI training as such a "new use." It alleges that thousands of recordings covered by the SRLA were included in licenses granted by Universal and Warner to Suno and Udio.
The court filing also contrasts the current licenses with the labels' June 2024 copyright complaints against the two AI companies. In those earlier suits, the labels alleged that the services were "replacing the work of human artists with massive quantities of AI created sounds ... that substantially dilute the royalty pools paid out to artists."
According to the AFM's amended complaint, settlements reached in late 2025 included compensation for past alleged infringement and licenses for catalog use on a retroactive and prospective basis. Those assertions are allegations in the union's filing, not findings by the court.
A labor issue alongside AI licensing
The case focuses on performer compensation and contractual disclosure rather than whether model training itself infringes copyright. That distinction matters because AI music licensing agreements can resolve rights-holder claims while leaving separate contractual obligations to performers contested.
For data and ML practitioners, disputes of this kind illustrate a recurring governance issue in training-data procurement: ownership of a catalog license does not necessarily settle compensation, consent, attribution, or reporting obligations for every contributor to the underlying works. Comparable disputes can make dataset provenance and rights metadata operational concerns, particularly where models are trained on commercially licensed media.
The amended complaint highlights Udio's reported "Starstruck" service. Digital Music News reported that the filing describes a workflow in which users select a particular artist and recording before generating output, with resulting recordings reportedly owned by participating rights holders rather than users. The AFM alleges this use further demonstrates a commercial exploitation of musicians' performances covered by the SRLA.
Key Points
- 1AFM's amended complaint contests whether label AI licenses trigger contractual performer compensation and disclosure obligations under the Sound Recording Labor Agreement.
- 2The filing substitutes Warner label entities after Warner Music Group disputed that its parent company was the proper contract defendant.
- 3Comparable training-data disputes make contributor-level rights metadata and licensing provenance important operational concerns for commercial generative media systems.
Scoring Rationale
The lawsuit directly concerns contractual compensation and disclosure for recordings licensed to major AI music training systems. It is a notable governance and labor dispute for teams building or procuring generative-media datasets, although the court has not ruled on the claims.
Sources
Primary source and supporting public references used for this report.
Practice interview problems based on real data
1,625 SQL & Python problems across 15 industry datasets — the exact type of data you work with.
Try 250 free problems
