PVLF 2025

Anthropic Wins Initial Round in AI Copyright Dispute Against Music Publishers

A U.S. judge denied music publishers' bid to block Anthropic from using lyrics for AI training, ruling no "irreparable harm." The copyright battle continues.
on Mar 26, 2025

A California federal judge has rejected music publishers' plea to enjoin AI firm Anthropic from using copyrighted lyrics to train its chatbot, Claude.

U.S. District Judge Eumi Lee determined the publishers, which include Universal Music Group (UMG), Concord, and ABKCO, did not prove "irreparable harm" and that their application was too broad. The publishers, which sued Anthropic in 2023 for purportedly copying at least 500 songs' lyrics without consent, were confident about their case in spite of the decision.

Anthropic embraced the ruling, labeling the publishers' request "disruptive and amorphous." The suit is one of a series of legal disputes regarding AI training methods, with other cases involving writers, news organizations, and visual artists.

Technology firms such as OpenAI, Microsoft, and Meta claim that training AIs is "fair use" under U.S. copyright law. Although fair use was anticipated to be the central issue of the case, Lee's ruling did not mention it directly. The judge also refused to identify a market for licensing AI training, stating that the fair use question is still open.

For the latest news, Visit: Frontlist