Sony Music Publishing, Warner Chappell Music, and affiliated publishers filed a lawsuit against AI company Anthropic, its CEO Dario Amodei, and co-founder Benjamin Mann on August 28. The suit alleges that Anthropic unlawfully used copyrighted songs and lyrics to train its Claude AI models, accusing the company of piracy through torrenting, web scraping, and unauthorized dataset use, according to medianama.com.
The complaint, filed in the US District Court for the Northern District of California, claims Anthropic downloaded millions of pirated books from sources like LibGen and PiLiMi, which included copyrighted lyrics and sheet music. It also alleges that the company scraped lyrics from websites such as MusixMatch and LyricFind. The publishers argue that Claude can reproduce copyrighted lyrics and generate derivative or competing works based on those materials.
The publishers describe the alleged conduct as one of the largest intellectual property thefts in history, seeking damages, an injunction against Anthropic, disclosure of the training data, and destruction of infringing copies. The case highlights ongoing legal challenges faced by AI firms regarding the use of copyrighted content in training datasets, a critical issue as AI-generated content grows in prominence.
The lawsuit remains pending in the Northern District of California, with the publishers pushing for court orders to halt the alleged infringement and require transparency from Anthropic about its data sources, per medianama.com.