A federal judge in Massachusetts has allowed major record labels to pursue a new Digital Millennium Copyright Act (DMCA) claim against Suno, adding a potentially important anti-circumvention issue to the wider fight over AI music training.
Photo: Giuseppe Di Maria / Pexels · Free to use
Court lets the DMCA claim move forward
According to Bloomberg Law, U.S. District Judge F. Dennis Saylor IV ruled that UMG Recordings and the other label plaintiffs had plausibly alleged a DMCA violation and could amend their complaint. The ruling does not establish that Suno committed stream ripping. It means the allegation is legally sufficient to proceed at this stage of the case.
The labels allege that Suno obtained copyrighted recordings from YouTube for its training dataset by using downloading tools that bypassed technological measures controlling access to the underlying media files. Court filings have identified tools including YT-DL and YT-DLP.
What “stream ripping” means in this case
Stream ripping generally refers to extracting a permanent media file from content made available primarily for streaming. In their proposed amended complaint, the labels say YouTube uses technical controls, including a changing or “rolling” cipher, to govern access to media files and prevent unrestricted downloading.
The labels contend that Suno used third-party tools to get around those controls and save recordings for use in model training. Suno has previously acknowledged that its training data included music available on the open internet and has argued that training generative AI on copyrighted works is protected by fair use. The new DMCA theory is distinct from that fair-use dispute.
Why the DMCA angle changes the case
Copyright infringement and anti-circumvention are related but legally different questions. Suno’s central fair-use argument addresses whether copying works to train an AI model can be lawful. A DMCA claim focuses instead on whether technological protection measures were bypassed in order to gain access to those works.
Photo: Brett Sayles / Pexels · Free to use
That distinction could matter well beyond Suno. If courts accept that particular scraping or downloading methods violate anti-circumvention law, AI developers may face legal exposure even before a court reaches the larger question of whether model training itself is transformative fair use.
Suno’s position
Suno had argued that the proposed DMCA claim should not be allowed to proceed. Earlier briefing challenged the labels’ theory that bypassing YouTube’s technical measures amounted to unlawful circumvention. The judge’s latest ruling rejects that attempt at the pleading stage and allows the labels to litigate the claim.
The underlying copyright case remains unresolved. Suno continues to defend the use of copyrighted music for AI training, while the labels maintain that the company copied protected recordings without permission and now allege that some of those copies were obtained through prohibited circumvention.
What happens next?
The amended complaint can now include the DMCA claim, which means discovery and later motions may examine both the technical process used to obtain recordings and the legal status of those access controls. The court will still have to decide the merits of the allegation, and Suno will have opportunities to contest the evidence and legal theory.
For the wider AI music industry, the case is becoming a test not just of what developers may use for training, but of how they acquire that material in the first place.
