Sony Music Publishing and Warner Chappell Music have filed a massive copyright lawsuit against Anthropic, its CEO Dario Amodei, and co-founder Benjamin Mann, accusing the AI company of orchestrating "one of the largest and most blatant ongoing thefts of intellectual property in history."

The complaint, filed in the Northern District of California on August 28, alleges that Anthropic used "tens of thousands" of copyrighted song lyrics to train Claude, its flagship AI assistant. The publishers claim the lyrics were obtained through two methods: torrenting from Library Genesis and Pirate Library Mirror, and scraping from lyric databases MusixMatch and LyricFind.

The lawsuit brings four counts: direct infringement by torrenting against all three defendants, contributory infringement by torrenting, direct infringement through scraping, and removal of copyright management information. The named works include iconic songs such as "Ain't No Mountain High Enough," "All I Want for Christmas Is You," and "Eye of the Tiger."

If willful infringement is proven, the publishers are seeking up to $150,000 per infringed work — which, given the "tens of thousands" of compositions alleged, could total billions of dollars. An additional $25,000 penalty applies for each instance of Anthropic allegedly removing copyright management information from the works.

The complaint describes a detailed forensic trail. The publishers allege that Anthropic's training data pipeline pulled lyrics from shadow library sites that host pirated books and other copyrighted material. The lawsuit cites specific BitTorrent swarm data linking Anthropic's infrastructure to downloads from these archives.

"This is not a case about fair use or transformative AI," the complaint states. "This is a case about theft — about a company that chose to steal rather than license, and then tried to hide the evidence."

Anthropic has not yet publicly commented on the lawsuit. The case could have far-reaching implications for how AI companies source training data, particularly for models that generate or discuss copyrighted creative works.

The lawsuit arrives amid a wave of copyright litigation against AI companies. The New York Times has sued OpenAI and Microsoft, the recording industry has sued multiple AI music generators, and visual artists have filed class actions against image generation companies. But the Sony-Warner case is notable for the specificity of its evidence — including alleged BitTorrent swarm data — and the sheer scale of potential damages.

Sources