Sony Music, Warner sue Anthropic, alleging a “brazen campaign” of intellectual property theft

Music Publishers File Copyright Lawsuit Against Anthropic

Sony Music Publishing, Warner Chappell, and other major music publishers have filed a lawsuit against AI lab Anthropic, accusing the company of systemic and widespread copyright infringement. The complaint, filed in the U.S. District Court for the Northern District of California, alleges that Anthropic engaged in a “brazen campaign” to illegally acquire copyrighted materials, including song lyrics and sheet music, to train its large language model, Claude. This action represents another significant legal challenge from the creative industries, intensifying the battle over the use of protected intellectual property in AI development.

The lawsuit builds on previous legal actions against the AI firm, specifically accusing Anthropic and its co-founders of obtaining training data through illegal torrenting and scraping. This case is distinct from prior suits due to its focus on the method of data acquisition. It follows the landmark Bartz v. Anthropic case, where a judge ordered Anthropic to pay $1.5 billion, establishing a precedent that while using copyrighted works for training may be permissible, acquiring them via piracy is not. An Anthropic spokesperson stated the company disagrees with the claims and intends to defend itself in court.

• Plaintiffs: Sony Music Publishing, Warner Chappell, and numerous other music publishers.
• Defendant: AI company Anthropic and co-founders Dario Amodei and Benjamin Mann.
• Core Allegation: Using illegally torrented and scraped copyrighted works, including lyrics and sheet music, to train the Claude AI model.
• Legal Precedent: Follows a $1.5 billion judgment against Anthropic in a previous case that differentiated between using and illegally acquiring copyrighted training data.

This lawsuit underscores a critical vulnerability for the AI industry: the provenance of training data. As legal challenges mount, the defense of 'fair use' is being tested, particularly when the data acquisition methods themselves are illegal. The outcome of this case could establish new standards for data sourcing and force AI developers to invest heavily in licensed or verifiably clean datasets. This shift could create substantial financial and operational hurdles for model development, potentially altering the competitive landscape and slowing the pace of innovation for firms unable to secure legally sound data pipelines.

The legal strategy against AI labs is shifting from a broad 'fair use' debate to a more targeted attack on the provenance of training data, creating a direct and potentially costly liability for companies that cannot verify the legal acquisition of their datasets.