Music Publishers Sue Anthropic Over Alleged Copyright Infringement

Major music publishers, including Sony Music Publishing and Warner Chappell Music, have sued artificial intelligence company Anthropic. The lawsuit, filed on August 28 in the U.S. District Court for the Northern District of California, alleges that Anthropic unlawfully obtained and used copyrighted songs to train its Claude AI models.

The publishers claim Anthropic acquired copyrighted material through illicit means, including pirate libraries like LibGen and PiLiMi, as well as by scraping lyrics from websites such as MusixMatch and LyricFind. The complaint also states that Anthropic utilized third-party datasets, including Books3, The Pile, and Common Crawl, which allegedly contain unauthorized works. Furthermore, the suit details how Anthropic digitized physical books by destroying the original copies.

According to the complaint, Anthropic's Claude AI is capable of reproducing copyrighted lyrics verbatim or near-verbatim. It is also alleged that the AI can generate derivative lyrics or new lyrics that compete with existing copyrighted works. The publishers are seeking damages, an injunction, disclosure of Anthropic's training data, and the destruction of allegedly infringing copies.

This legal action follows previous findings where Anthropic was identified as having downloaded large quantities of pirated books. While Anthropic denies directly using material from LibGen and PiLiMi for its commercial models, the publishers contend that commercial models have been influenced by other Anthropic models trained on data derived from these sources. The lawsuit covers numerous well-known compositions, such as "Ain’t No Mountain High Enough" and "All I Want for Christmas is You," though the allegations have not yet been proven in court.