Key Points
- Music publishers led by Concord and Universal sue Anthropic for alleged illegal downloading of over 20,000 songs.
- The songs, including sheet music, lyrics, and compositions, were reportedly used to train Anthropic’s Claude chatbot.
- Publishers claim potential damages could exceed $3 billion, making it one of the largest non‑class action copyright cases in the U.S.
- The lawsuit cites iconic tracks by The Rolling Stones, Neil Diamond, and Elton John as examples.
- Reference is made to the prior Bartz v. Anthropic case, which resulted in a $1.5 billion award for writers.
- The earlier case established that training AI on copyrighted material is legal, but acquiring that material via piracy is not.
- Anthropic is described as being valued at around $350 billion despite the allegations.
- The same legal team that handled the Bartz case is representing the publishers in this new suit.
Background of the Lawsuit
A group of music publishers, headed by Concord Music Group and Universal Music Group, filed a lawsuit against AI company Anthropic. According to the filing, Anthropic allegedly downloaded more than 20,000 copyrighted songs—covering sheet music, lyrics, and full compositions—through illegal means. The plaintiffs assert that these works were subsequently fed into Anthropic’s chatbot, Claude, for the purpose of training its artificial‑intelligence models.
Claims and Potential Damages
The publishers contend that the unauthorized use of the songs could result in damages that exceed $3 billion. If accurate, such a figure would rank the case among the largest non‑class action copyright lawsuits in United States history. The complaint lists iconic tracks by artists such as The Rolling Stones, Neil Diamond, and Elton John as examples of the works at issue.
Industry Context and Prior Litigation
The lawsuit references a previous legal battle, Bartz v. Anthropic, in which a jury awarded $1.5 billion to affected writers after finding that Anthropic had similarly downloaded copyrighted material for training purposes. That case established a precedent that, while training AI models on copyrighted content may be permissible, acquiring the content through piracy is not. The Bartz settlement stipulated that roughly 500,000 authors would receive about $3,000 per work.
Anthropic’s Position and Business Valuation
The filing criticizes Anthropic’s public portrayal as an “AI safety and research” company, arguing that its business model relies on pirated content. The plaintiffs note that Anthropic’s valuation is reported to be around $350 billion, underscoring the scale of the company implicated in the alleged infringement.
Legal Implications and Next Steps
The suit was brought by the same legal team that represented the plaintiffs in the Bartz case. The plaintiffs claim that discovery in the earlier lawsuit revealed the illegal downloading activity now central to the current complaint. The outcome of this case could clarify the legal boundaries between permissible AI training and unlawful acquisition of copyrighted works, potentially influencing how AI developers source training data in the future.
Source: engadget.com