Sony, Warner Music Sue Anthropic Over AI Training Data

Sony Music Publishing and Warner Chappell Music have initiated a copyright infringement lawsuit against artificial intelligence firm Anthropic. The complaint, filed in the US District Court for the Northern District of California, alleges a “brazen campaign of illegally torrenting, scraping, and downloading” copyrighted works.

These tens of thousands of music compositions, lyrics, and sheet music were purportedly used to train Anthropic’s AI models, including its Claude system. Anthropic co-founders Dario Amodei and Benjamin Mann are also named as defendants in this significant legal challenge.

Core Allegations Against Anthropic

• Anthropic allegedly torrented, scraped, and downloaded tens of thousands of copyrighted music compositions, lyrics, and sheet music.
• These materials were used to train Anthropic’s AI models, specifically its Claude system.
• Hundreds of song lyrics from prominent artists like Michael Jackson, Taylor Swift, and The Beatles were utilized without authorization.

This lawsuit represents the latest escalation in a series of legal actions by various copyright holders, including authors, publishers, music labels, and news organizations, against technology corporations leveraging creative works for AI system training.

Financial Context and Prior Legal Battles

• Universal Music Group (UMG.AS) sued Anthropic in 2023 for similar allegations; that lawsuit remains ongoing.
• Anthropic previously settled a class-action lawsuit with a group of authors for $1.5 billion last year.
• Plaintiffs cite Anthropic’s reported $2-trillion valuation, arguing the prior settlement was insufficient to deter infringing conduct.

Sony and Warner argue that Anthropic views such settlements as merely a “cost of doing business,” suggesting the company’s entire business model is predicated on copyright theft. They contend the prior settlement failed to deter continued infringement.

This new legal action significantly intensifies the ongoing conflict between developers of generative AI and traditional media publishers over intellectual property rights.