In a groundbreaking move, three major music publishers—Universal Music Group, Concord Music Group, and ABKCO Music—have filed a lawsuit against AI company Anthropic, alleging that its AI models infringe on copyrights related to song lyrics. The lawsuit, filed in the Nashville division of the U.S. District Court for the Middle District of Tennessee, represents a fresh approach by the music industry to combat AI companies that employ copyrighted materials in training their models.
At the heart of the legal action is Claude, Anthropic’s AI model, which the publishers claim has been trained on content from various sources across the internet, including lyrics to numerous songs for which the plaintiffs hold publishing rights. Consequently, Anthropic’s AI chatbot can provide lyrics to songs like “Sweet Home Alabama” or “Gimme Shelter” upon request.
The lawsuit states, “There are already a number of music lyrics aggregators and websites that serve this same function, but those sites have properly licensed Publishers’ copyrighted works to provide this service.”
Beyond simply generating lyrics upon direct request, the suit alleges that Anthropic’s chatbot replicates lyrics from copyrighted songs even when responding to indirect prompts, such as creating a song on a related topic or composing poetry or short fiction in the style of a specific artist, all without acknowledging any rights holders. As part of their legal case, the companies have attached an exhibit containing 500 songs, spanning artists from the Beach Boys to Ariana Grande, for which they claim copyright infringement.
Furthermore, the publishers argue that Anthropic has benefited financially by licensing and offering access to chatbots powered by Claude and the API integrated into software developed by Anthropic’s commercial clients. The lawsuit emphasizes, “One of the reasons that Anthropic’s AI models are so popular and valuable is because of the substantial underlying text corpus that includes Publishers’ copyrighted lyrics.”
This legal action, targeting an AI company that has developed a Language Model (LLM), aligns with the music industry’s growing focus on combating generative AI companies that may encroach upon their copyrights. Anthropic, known for promoting a responsible approach to building its AI model with a set of guiding principles aimed at minimizing harm, has come under scrutiny, illustrating the music industry’s determination to protect its intellectual property.
The move is consistent with the Human Artistry Campaign, an industry-wide initiative launched by the RIAA and major labels like UMG earlier this year, which emphasizes responsible AI use in the music industry. Before this lawsuit, the efforts mainly concentrated on how generative music tools train their models, with UMG urging streaming platforms like Spotify and Apple Music to prevent companies from using their catalogs for AI model training.
In September, UMG announced a partnership with streaming service Deezer to promote music by artists with a specified number of monthly listeners and to reduce the presence of what it calls “non-artist noise audio.”
As of now, Anthropic has not responded to requests for comment regarding the lawsuit.
