Musicbusinessworldwide.com
BMG Rights Management has launched a legal challenge against Anthropic, escalating the music industry’s ongoing battle over how artificial intelligence systems are trained.
Filed in a federal court in California on March 17, the complaint accuses the company behind the Claude chatbot of improperly using hundreds of copyrighted songs to build its large language models.
According to the filing, nearly 500 compositions are at risk; the tracks range from classics like “What a Wonderful World” to modern hits including “Uptown Funk” and “Kryptonite.”
At the heart of the case is the claim that Anthropic not only ingested protected material without authorization, but also reproduced portions of lyrics in its AI-generated outputs. The publisher further alleges that the company sourced training data from unauthorized online repositories and stripped identifying copyright information before incorporating the material into its systems.
BMG argues that these practices undermine both legal standards and the livelihoods of songwriters. The company maintains that technological innovation does not exempt firms from securing proper licenses, and that the alleged actions have caused direct financial and creative harm to rights holders.
This is far from an isolated dispute. Universal Music Publishing Group, alongside Concord and ABKCO, has already taken similar action, with a separate case covering tens of thousands of songs and potentially billions in damages. A judge previously allowed that case to proceed, signaling that courts are increasingly willing to scrutinize how AI developers source their training data.
BMG, which controls rights to millions of compositions, says it remains open to the creative potential of AI, but only within a framework that guarantees consent, credit, and compensation.
About the Author
