The Sony Warner Anthropic copyright lawsuit expanded the music industry’s legal confrontation with artificial intelligence on 28 August, when Sony Music Publishing and Warner Chappell Music filed a Variety-reported 48-page complaint in the US District Court for the Northern District of California, naming Anthropic and its cofounders Dario Amodei and Benjamin Mann as defendants.
What the Sony Warner Anthropic copyright lawsuit alleges
The complaint accuses Anthropic of conducting what the publishers call a ‘brazen campaign of illegally torrenting, scraping, and downloading copyrighted works on a massive scale in order to develop, operate, and reap enormous profits’ from the Claude series of AI models. The publishing companies say Anthropic collected ‘thousands upon thousands’ of copyrighted songs, citing works including ‘Eye of the Tiger,’ Marvin Gaye’s ‘Ain’t No Mountain High Enough,’ Mariah Carey’s ‘All I Want for Christmas Is You,’ and Taylor Swift’s ‘Paper Rings.’
The alleged methods of acquisition are detailed in the filing. According to the complaint, Anthropic used two digital archives, Library Genesis and Pirate Library Mirror, to obtain books containing lyrics and sheet music for compositions including ‘Livin’ On a Prayer,’ ‘September,’ ‘Great Balls of Fire,’ ‘Ramblin’ Man,’ and ‘Hallelujah.’ This is consistent with a June 2025 ruling in which a judge found that Anthropic had downloaded over 7 million pirated books to train Claude.
The publishers further allege that training Claude on this material allows the model to generate identical or near-identical reproductions of copyrighted lyrics in response to user prompts, and that AI-generated output will ultimately compete directly with human-authored songs.
Damages sought and the copyright management information claim
Sony Music Publishing and Warner Chappell are seeking statutory damages of up to $150,000 for each copyrighted composition that Anthropic used to train Claude. According to The Verge, the companies are also seeking up to $25,000 for each instance in which identifiable copyright management information was stripped from the works. That second head of damages, not widely discussed in initial coverage, matters because the removal of rights metadata is treated separately under US copyright law and can substantially increase the aggregate exposure Anthropic faces.
The publishers have requested a jury trial. In the complaint, they wrote: ‘Even the most revolutionary of technologies must develop within the bounds of the law, and Anthropic’s Claude models are no different.’
Context: a growing wave of AI copyright claims
The action arrives at a delicate moment for Anthropic. According to Music Business Worldwide, citing a Forbes report, Anthropic carried a projected valuation of $2 trillion ahead of an anticipated IPO, making the financial stakes of cumulative litigation all the more consequential for the company and its investors. The publishers are evidently aware of those economics: the complaint is framed not merely as a rights enforcement action but as a challenge to a business model that, in their characterisation, is built in part on unlicensed creative work.
Anthropic is not alone in facing this pressure. The AI industry has been targeted by a series of copyright claims since large language models began processing internet-scale data. Anthropic itself previously agreed to pay over $1.5 billion to authors to settle a separate class-action lawsuit related to pirated works. OpenAI has faced litigation from The New York Times, Encyclopedia Britannica and others.
According to the Economic Times, the complaint was lodged in the Northern District of California on 28 August, the same court that has handled several of the most prominent AI copyright actions to date. The case now joins a queue of proceedings that will, collectively, begin to define whether training large language models on scraped copyrighted content constitutes infringement, and at what scale damages are calculated when it does.