Music Publishers Escalate AI Copyright Battle with Anthropic
Music publishers are intensifying their legal challenge against Anthropic, the artificial intelligence firm behind the Claude chatbot, alleging widespread copyright infringement related to the training of its large language model. The publishers are seeking a ruling that Anthropic’s employ of song lyrics was not protected under fair use doctrine, arguing it was a commercial exploitation of copyrighted works.
Copyright Claims and Legal Action
Universal Music Group (UMG), Concord Music Group, and ABKCO have filed a motion for partial summary judgment in a California federal court, seeking an immediate victory on the central question of fair use. The lawsuit, initially filed in October 2023, claims Anthropic copied lyrics from at least 499 copyrighted musical works without permission [Reuters]. A separate, larger lawsuit filed in January 2026, covers over 20,000 songs and seeks over $3 billion in statutory damages [Music Business Worldwide].
Allegations of Data Scraping and Copyright Removal
Beyond UMG, Concord, and ABKCO, music publisher BMG has also filed a copyright infringement lawsuit against Anthropic in a US federal court [Medianama]. BMG alleges Anthropic engaged in a multi-stage pattern of infringement, including scraping lyrics from online sources and downloading pirated material from “shadow libraries” like LibGen and PiLiMi using BitTorrent. The complaint further alleges Anthropic removed copyright management information (CMI) during data processing and stored the materials for indefinite use in training its models [Medianama].
The Fair Use Debate
The core of the dispute revolves around the application of “fair use” doctrine. Publishers argue that Anthropic’s actions were “quintessential infringement” – not fair use – because the AI company is a $380 billion entity that profited from the unauthorized use of copyrighted material [Billboard]. They contend that the doctrine was not intended to shield “unabashedly commercial” exploitation of copyrighted works that directly compete with the originals [Billboard].
Broader Implications for the AI Industry
This case is one of many copyright challenges facing AI companies. Anthropic, along with Meta, OpenAI, and Google, are all embroiled in similar lawsuits over the use of copyrighted material in training their AI models. A $1.5 billion settlement was reached in a separate case involving authors alleging the use of pirated books in AI training [Medianama]. The outcome of these cases could have significant implications for the future of AI development and the protection of intellectual property.