A federal judge in Massachusetts has refused to dismiss copyright claims against Suno, the AI music generator, in a putative class action brought by independent recording artists.
What the judge decided
US District Judge F. Dennis Saylor IV denied Suno's motion to dismiss claims under the Copyright Act and the Digital Millennium Copyright Act. He granted dismissal only of a claim under the Tennessee Consumer Protection Act.
The reasoning is the part that matters. Suno argued the case should fail because the plaintiffs had not pointed to a single generated track that is substantially similar to one of their songs. The judge rejected that as a categorical requirement, holding that plaintiffs pleaded enough to proceed even without producing the derivative work at this stage.
Why that is significant
Identifying a specific infringing output is the hardest evidence to produce against a generative model, because outputs are produced on demand and are never identical twice. If that were required before discovery, most claims of this kind would end at the first hearing.
The plaintiffs also allege their recordings were downloaded from YouTube, which is what the DMCA claim rests on. That claim moving forward separates the question of how training data was obtained from the question of what the model produces.
What happens next
Discovery. That is the practical consequence of surviving a motion to dismiss, and it is where cases like this become expensive and revealing, because it forces disclosure of what went into the training set and how.
What it means for people using these tools
Nothing changes today for anyone generating music with Suno. But if you are building commercial work on the output of a generative music tool, the licensing terms you rely on are now being tested in court rather than accepted. Keep records of what you generated and when, and read the terms your provider actually offers on commercial use.