Music publishers file sweeping copyright suit against Anthropic over alleged piracy
Sony Music Publishing, Warner Chappell, and other publishers sued Anthropic and its co-founders on Friday, accusing the AI lab of illegally torrenting and scraping copyrighted works to train its Claude model. The lawsuit, filed in California, builds on earlier cases and specifically alleges flagrant piracy of books containing lyrics and sheet music. Anthropic had not responded to requests for comment at the time of reporting.
The lawsuit, filed in the Northern District of California, names Anthropic’s co-founders personally and follows a January case brought by Concord Music and Universal Music Group, with overlapping legal counsel. It also builds on the Bartz v. Anthropic ruling, which ordered a $1.5 billion payout after finding that while using copyrighted works was permissible, obtaining them through piracy was not. Publishers now allege Anthropic used illegal torrenting to amass millions of books containing lyrics and sheet music, calling the practice “flagrant piracy.” Anthropic had not commented by publication time.
This action widens the legal pressure on AI training practices, moving beyond authors to the music industry. The earlier Bartz case established a key distinction between lawful use and unlawful acquisition, which this suit leverages. The publishers’ broad scope—covering lyrics and sheet music—suggests they aim to set a precedent for how AI firms source training data, potentially forcing greater transparency in data procurement across the sector.
This lawsuit could reshape how AI companies obtain training data, especially for creative industries. If successful, it may force Anthropic and others to license content or prove lawful sourcing, raising costs and slowing model development. Musicians and publishers could gain stronger protections, while smaller AI startups may struggle to compete. The case also signals that courts are scrutinizing not just fair use, but the methods of acquisition, potentially setting standards that affect all generative AI users and content creators.