Suno Moves to Keep Size of Its AI Training Data Sealed in UMG and Sony’s Copyright Case, Citing ‘Competitive Harm’

Why It Matters

Sealing the training‑data size protects Suno’s proprietary development strategy while highlighting the tension between copyright transparency and competitive secrecy in the fast‑growing AI music market.

Key Takeaways

• Suno wants only the total file count sealed, not specific tracks
• Labels allege Suno used millions of copyrighted recordings
• Court previously allowed similar data impoundments in AI cases
• Warner Music settled with Suno; UMG and Sony still litigating

Pulse Analysis

The dispute centers on Suno's request to keep the precise volume of audio files used to train its AI music generator confidential. While the broader lawsuit by Universal Music Group and Sony Music alleges that Suno incorporated millions of copyrighted recordings, Suno argues that revealing the exact count would enable competitors to reverse‑engineer its model, benchmark performance, and erode its market advantage. By limiting the seal to a single numeric figure, Suno seeks to protect strategic data without hiding the identities of the works themselves, a nuance that courts have previously recognized in AI‑related copyright fights.

This filing underscores a growing legal crossroads where copyright enforcement meets the proprietary nature of AI training data. Plaintiffs argue that the size of the training corpus is central to proving infringement, yet defendants contend that such metrics are trade secrets essential for competitive positioning. Recent rulings, including those in *The New York Times v. Microsoft* and *Concord Music Group v. Anthropic*, have set precedents for sealing sensitive AI data, suggesting courts may favor protecting business interests over full disclosure, especially when the information does not directly identify infringing works.

Looking ahead, the outcome could shape how the music industry negotiates licensing with AI developers. Warner Music's 2025 settlement with Suno demonstrates a pathway for collaboration, while UMG and Sony's stalled talks reflect lingering mistrust. If the court upholds Suno's seal, it may encourage more AI firms to argue for data confidentiality, potentially limiting transparency in future copyright cases. Conversely, a decision to unseal could pressure AI companies to adopt more open data practices, influencing licensing standards and the competitive dynamics of AI‑generated music.