A federal judge refused on August 18, 2026, to add 61,026 recordings to the labels’ existing copyright case against Suno. The refusal was not a ruling that those additional infringement claims were weak or invalid.
“Denied without prejudice” leaves the door open. UMG Recordings and Sony Music can still pursue claims involving those recordings, but the court decided the original case was the wrong vehicle for dropping tens of thousands of new works into discovery.
The original lawsuit identified 560 recordings as an illustrative sample of material allegedly copied for AI training. Months of discovery later produced a much larger proposed set, while the separate fight over Suno’s admitted use of YouTube audio added another contested issue to an already dense case.
Timing hurt the labels here. They sought to add the 61,026 works roughly one month before the scheduled close of fact discovery. The discovery deadline was later extended to September 30, 2026, already more than a year and a half beyond the original deadline identified by the court.
Adding a track is not merely typing another title into a complaint. Each asserted recording can bring questions about ownership, registration, whether it appeared in the training material, what copying allegedly occurred, possible defenses, and damages. Multiplying those disputes by 61,026 could turn a case nearing the end of fact discovery into a discovery project.
Judge F. Dennis Saylor IV said the magnitude of alleged infringement was not a defense for Suno. His concern was procedural. Expanding the case at that stage would likely prejudice Suno by delaying resolution for an indefinite period.
Suno’s fair-use defense sat at the center of that concern. The court described summary judgment on fair use as likely to resolve the predominant issue in the case and said Suno was entitled to a timely decision. Loading 61,026 additional works into the same proceeding risked pushing that decision farther away.
Another route would have allowed amendment while severing and staying the 61,026 new claims. The labels proposed a form of bifurcation, letting the original fair-use issue move toward summary judgment while discovery concerning the expanded catalog waited. The judge saw both approaches as possible but preferred parallel cases as the cleaner balance.
“Without prejudice” therefore carries real weight. It means the August order controlled where and when the claims could proceed, not whether the labels ultimately possess valid infringement claims. Sony had already demonstrated the practical version of this move in its Udio litigation by filing a second case after a judge refused to add more than 30,000 recordings to the older lawsuit.
Parallel litigation can still become coordinated later. The Suno order contemplated different schedules, a possible stay, and even future consolidation in whole or in part. None of those possibilities required the original case to absorb the new works immediately.
Rule 54(b) deals with final judgments when a case contains multiple claims or parties and only some have been resolved. Courts have wrestled with its boundaries for decades, which is why an academic critique of partial final judgments describes the rule as a persistent source of appellate difficulty. The judge did not present that complication as fatal, only as one cost of managing such a large expansion.
The court wanted the fair-use question in the 560-work case to reach a decision without reopening discovery across 61,026 new works. If the labels bring the 61,026 recordings separately, those claims can be preserved while the older case keeps its existing procedural shape.
A later fair-use ruling could influence how much work a parallel case needs. If the ruling resolves a common training issue, the parties and court may have a clearer map for handling thousands of additional recordings. If it does not, the extra works remain available for their own litigation without having delayed the first case before the central issue was decided.
“Denied without prejudice” leaves the door open. UMG Recordings and Sony Music can still pursue claims involving those recordings, but the court decided the original case was the wrong vehicle for dropping tens of thousands of new works into discovery.
The original lawsuit identified 560 recordings as an illustrative sample of material allegedly copied for AI training. Months of discovery later produced a much larger proposed set, while the separate fight over Suno’s admitted use of YouTube audio added another contested issue to an already dense case.
The court focused on delay rather than claim size
Federal Rule of Civil Procedure 15 tells courts to permit amendments to pleadings freely when justice requires it. Permission is not automatic. Judges can weigh undue delay, prejudice, futility, repeated amendment problems, bad faith, and the practical cost of making an existing lawsuit harder to finish.Timing hurt the labels here. They sought to add the 61,026 works roughly one month before the scheduled close of fact discovery. The discovery deadline was later extended to September 30, 2026, already more than a year and a half beyond the original deadline identified by the court.
Adding a track is not merely typing another title into a complaint. Each asserted recording can bring questions about ownership, registration, whether it appeared in the training material, what copying allegedly occurred, possible defenses, and damages. Multiplying those disputes by 61,026 could turn a case nearing the end of fact discovery into a discovery project.
Judge F. Dennis Saylor IV said the magnitude of alleged infringement was not a defense for Suno. His concern was procedural. Expanding the case at that stage would likely prejudice Suno by delaying resolution for an indefinite period.
Suno’s fair-use defense sat at the center of that concern. The court described summary judgment on fair use as likely to resolve the predominant issue in the case and said Suno was entitled to a timely decision. Loading 61,026 additional works into the same proceeding risked pushing that decision farther away.
A separate lawsuit keeps the extra works alive
The order sketched two workable routes instead of treating the extra claims as dead. One route was the one the court chose. Deny amendment, then leave the labels free to file another lawsuit. A new Massachusetts case would likely be treated as related, could land before the same judge, and could run on a different discovery schedule.Another route would have allowed amendment while severing and staying the 61,026 new claims. The labels proposed a form of bifurcation, letting the original fair-use issue move toward summary judgment while discovery concerning the expanded catalog waited. The judge saw both approaches as possible but preferred parallel cases as the cleaner balance.
“Without prejudice” therefore carries real weight. It means the August order controlled where and when the claims could proceed, not whether the labels ultimately possess valid infringement claims. Sony had already demonstrated the practical version of this move in its Udio litigation by filing a second case after a judge refused to add more than 30,000 recordings to the older lawsuit.
Parallel litigation can still become coordinated later. The Suno order contemplated different schedules, a possible stay, and even future consolidation in whole or in part. None of those possibilities required the original case to absorb the new works immediately.
Splitting the claims creates procedural friction
Keeping the original 560-work case moving is cleaner, but it is not frictionless. If claims move on separate tracks while related issues remain unresolved elsewhere, appeal timing can become awkward. The judge noted that a separate judgment under Rule 54(b) might be needed in one possible bifurcated setup to permit a timely appeal.Rule 54(b) deals with final judgments when a case contains multiple claims or parties and only some have been resolved. Courts have wrestled with its boundaries for decades, which is why an academic critique of partial final judgments describes the rule as a persistent source of appellate difficulty. The judge did not present that complication as fatal, only as one cost of managing such a large expansion.
The court wanted the fair-use question in the 560-work case to reach a decision without reopening discovery across 61,026 new works. If the labels bring the 61,026 recordings separately, those claims can be preserved while the older case keeps its existing procedural shape.
A later fair-use ruling could influence how much work a parallel case needs. If the ruling resolves a common training issue, the parties and court may have a clearer map for handling thousands of additional recordings. If it does not, the extra works remain available for their own litigation without having delayed the first case before the central issue was decided.
