level: business
sony music publishing, warner chappell, and other music publishers sued anthropic and its co-founders in federal court in california. the lawsuit claims anthropic ran a brazen campaign of illegally torrenting, scraping, and downloading copyrighted works. the publishers say anthropic used thousands of songs and lyrics to train its claude model without permission. anthropic said it disagrees with the claims and will defend itself in court. this is the latest in a series of copyright lawsuits against ai companies.
the complaint alleges anthropic used illegal torrenting to obtain millions of books containing lyrics and sheet music. this follows a january case by concord and universal music group, and the bartz v. anthropic case where a judge ordered anthropic to pay $1.5 billion. in bartz, the court ruled using copyrighted works for training was legal, but acquiring them through piracy was not. the new lawsuit builds on that ruling, accusing anthropic of flagrant piracy on a large scale.
the case highlights growing tension between ai developers and rights holders. music publishers argue that training on pirated material gives claude an unfair advantage and harms songwriters. anthropic has faced multiple lawsuits over training data, and this one adds music publishers to the list. the outcome could set rules for how ai companies source training data. for ai practitioners, it signals that data provenance and licensing are becoming critical legal risks.
why it matters: this case could force ai companies to prove their training data is legally sourced, affecting how models are built and deployed.