UMG and Sony Hit Suno With New Copyright Lawsuit Over V6 AI Music Model
Universal, Sony and Warner Music originally sued Suno in 2024 over alleged unauthorized use of copyrighted recordings. Warner later reached a licensing agreement with Suno, while BMG and Believe have also partnered with the AI company.
Universal Music Group and Sony Music Entertainment have launched a fresh legal challenge against AI music company Suno, arguing that its newly released v6 models remain connected to earlier systems allegedly trained on copyrighted recordings without permission.
The new lawsuit escalates a broader battle over how generative AI companies can use music to develop models and whether licensing should be required when copyrighted recordings become part of the training process.
Sony and UMG Allege 60,202 Recordings Were Copied
According to the complaint filed in the U.S. District Court for the District of Massachusetts, the two music giants identify 60,202 sound recordings that they allege Suno copied during its earlier training activities.
The labels say that number represents only a portion of the recordings they believe were infringed.
The complaint also focuses heavily on Suno’s new v6 models. Sony and UMG argue that although Suno has partnered with music companies including Warner Music Group, BMG and Believe, the new technology allegedly retains connections to previous models.
The labels point to Suno’s description of v6 as being trained partly on user interactions, creations, and preference signals. They argue that these interactions were generated through earlier Suno models.
Their central argument is that training a new model with outputs and information derived from allegedly infringing models does not eliminate the underlying copyright concerns.
Labels Target Suno’s AI Training Process
The Suno lawsuit describes v6 as being connected to what the labels call a continuing chain of unauthorized use.
Sony and UMG allege that Suno’s earlier models were trained using copyrighted recordings without authorization and that information generated by those systems subsequently contributed to the development of v6.
The labels therefore contend that the latest model cannot simply be treated as a clean break from previous versions.
The dispute highlights a major unresolved question in the AI music industry: Can an AI model be considered independently trained if its development incorporates outputs, preferences, or knowledge derived from an earlier model whose training data is disputed?
That question could have implications far beyond Suno as music companies and AI developers negotiate licensing agreements and battle over copyright law.
Suno Rejects the Allegations
Suno has strongly disputed the new allegations, describing the claims as fundamentally flawed both factually and legally.
The company said v6 was developed using licensed content from its partners, community interactions and creations, and accumulated technical knowledge from its team.
Suno also defended the purpose of its platform, arguing that its technology allows users to create original music rather than reproduce existing recordings.
The company said that when users reference an artist, its system translates the request into musical characteristics rather than allowing users to directly imitate another artist’s work.
The response comes as Suno attempts to build closer relationships with the traditional music industry through licensing agreements.
AI Music Copyright Battle Intensifies
The latest lawsuit is part of a rapidly expanding legal fight over AI-generated music.
Universal, Sony and Warner Music originally sued Suno in 2024 over alleged unauthorized use of copyrighted recordings. Warner later reached a licensing agreement with Suno, while BMG and Believe have also partnered with the AI company.
Sony and UMG’s latest complaint puts those agreements into the wider copyright debate, arguing that licensing markets show music companies and AI developers can negotiate permission for training uses.
Suno, meanwhile, has maintained that its technology should be protected under existing copyright principles and has challenged aspects of the labels’ claims.
The case could ultimately force courts to examine how copyright applies not only to the original datasets used to train AI systems, but also to synthetic outputs, model-to-model learning and subsequent generations of AI technology.
As generative music becomes increasingly sophisticated, the Suno dispute is emerging as another major test of where the music industry’s copyright protections end and where AI innovation begins.
