The new lawsuit against Anthropic signals an escalating confrontation between AI companies and the music industry, with billions of dollars—and the future rules of AI training—now hanging on how courts interpret copyright and fair use. (Source: Image by RR)

Music Publishers Claim Anthropic Used Copyrighted Songs without Permission

Artificial intelligence startup Anthropic is facing a new high-stakes lawsuit from major music publishers, including Universal Music Group, Concord, and ABKCO, over allegations that it improperly used copyrighted songs to train its chatbot, Claude. Filed in federal court in California, the suit claims Anthropic pirated lyrics and sheet music from more than 700 songs, with alleged infringements spanning over 20,000 works in total. The publishers argue the misuse constitutes one of the largest non-class-action copyright cases ever filed in the United States.

According to the complaint, Anthropic allegedly used copyrighted lyrics and musical compositions from artists such as The Rolling Stones, Neil Diamond and Elton John without permission. The publishers claim statutory damages could exceed $3 billion, asserting that Anthropic’s actions went far beyond incidental exposure and instead involved systematic copying of protected works. Anthropic, as noted in reuters.com, did not immediately respond to requests for comment and has denied similar allegations in prior cases.

The lawsuit builds on an earlier 2023 case brought by the same publishers, as well as a separate, closely watched legal battle involving book authors. In that case, Anthropic agreed to pay $1.5 billion to settle claims related to pirated books used in AI training. While a judge ruled that fair use protected certain aspects of AI training, he also suggested Anthropic could have faced up to $1 trillion in damages for unauthorized copying—highlighting the enormous legal risks surrounding AI data practices.

This latest case underscores the intensifying legal pressure facing AI companies as copyright holders across music, publishing and visual arts challenge how generative models are trained. While tech firms argue that training AI systems constitutes fair use, rights holders counter that large-scale ingestion of copyrighted works without licensing undermines creative industries. The outcome of the Anthropic case could set important precedents for how AI companies source data—and how much they may ultimately have to pay for it.

read more at reuters.com