Sony Sues Udio Over 30,000+ Songs in AI Copyright Fight
24 Jul 2026
Sony Music Entertainment has filed a new lawsuit in a New York court accusing AI music generator Udio of infringing the copyright of more than 30,000 of its songs — a filing that lands amid a fractured legal landscape where two of Sony's major-label peers have already settled with the same company.
What happened
The lawsuit, filed Monday, alleges Udio infringed copyrighted recordings from artists including Elvis Presley, Beyoncé, Harry Styles, Britney Spears, and Johnny Cash. Sony is seeking statutory damages of up to $150,000 per infringed work — a figure that, applied across 30,000+ songs, points to potentially enormous financial exposure for Udio, though the total amount sought is not calculated in available filings.
Central to Sony's claim is an allegation that Udio itself admitted its generative AI models were "constructed by showing the program a vast amount of different kinds of sound recordings," including material sourced from YouTube.
The backstory — and a key conflict
This isn't the first time Sony has gone after Udio. In 2024, Sony, Universal Music Group, and Warner Records jointly sued Udio and Suno, another AI music generator, over similar copyright concerns. A judge later rejected Sony's motion to expand that original case to include the same 30,000+ songs now named in the new suit, capping the case's scope at 333 works.
Sources do not clarify whether Monday's new filing is a separate lawsuit from that rejected motion or an attempt to refile the same claims through a new venue — an ambiguity worth watching as the case develops.
Meanwhile, UMG and Warner Music Group have taken a different path entirely: both settled with Udio and are now reportedly partnering with the company. The terms of those settlements haven't been disclosed, and Udio has not yet commented publicly on Sony's new lawsuit.
Why founders should care
For AI founders building products on scraped or unlicensed training data — particularly in music, but plausibly in other creative domains — this case is likely a signal worth heeding, not just a Sony-Udio dispute:
- Litigation risk from major rights holders may be rising, not falling. Sony's decision to continue pursuing Udio in court, even after UMG and Warner settled, suggests labels aren't uniformly moving toward negotiated resolutions. Founders relying on similar training approaches should probably assume litigation is a realistic possibility, not a tail risk.
- Statutory damages could be severe. The $150,000-per-work figure, if applied at scale, suggests that copyright exposure for AI training practices could plausibly reach into the hundreds of millions for companies operating at Udio's scale — a number startups should factor into legal risk modeling early, not after a suit is filed.
- Scraped platforms as training sources may carry outsized legal risk. Sony's claim that Udio trained on YouTube-sourced recordings hints that using publicly accessible but copyrighted content — even without direct licensing — may not shield companies from liability.
- Licensing deals may be a more durable strategy than litigation avoidance. UMG and Warner's settlements and subsequent partnerships with Udio suggest that proactive licensing negotiations, however costly upfront, could reduce long-tail legal exposure more effectively than betting on favorable court outcomes.
- Label-by-label risk tolerance varies. Sony's continued litigation stance versus UMG and Warner's settlements indicates that AI music startups likely need label-specific strategies rather than a one-size-fits-all licensing or defense approach.
What's still unclear
Several important details remain unresolved in current reporting: the precise relationship between the new 30,000-song lawsuit and the rejected 2024 motion, the financial terms of UMG's and Warner's settlements, Udio's response to the new filing, and the current status of the original 333-work case. The outcome of this suit — and whether it pushes Udio toward a Sony settlement similar to its peers — could help define legal norms for AI training data across the music industry, with implications for adjacent creative-AI sectors watching closely.