A federal judge appears ready to dismiss most of the copyright infringement lawsuit filed by music publishers against AI company Anthropic, while allowing claims about profits from alleged infringement to proceed.

During a 64-minute Zoom hearing on December 19th, the judge indicated she would likely dismiss two of three claims for being too general, specifically those regarding Anthropic's knowledge of alleged infringement. However, publishers will have the opportunity to amend and refile the dismissed claims.

The lawsuit alleges that Anthropic infringed upon compositions while training its Claude AI chatbot and through the chatbot's outputs. Publishers previously argued they had effectively demonstrated Anthropic's "direct financial benefit from infringing activity" and intentional removal of copyright management information.

This development represents another setback for rightsholders in AI-related litigation. While AI technology continues rapid advancement, legal cases concerning training data infringement are moving slowly. The Anthropic case isn't expected to go to trial until at least 2026.

Meanwhile, regulatory battles over AI training laws continue globally:

• The UK is considering legislation to allow tech firms to train on protected materials
• The EU continues to debate specifics of AI Act enforcement
• Multiple countries are developing frameworks for AI training data usage

The publisher plaintiffs have not yet publicly commented on these developments. The case highlights the ongoing tension between AI development and intellectual property rights protection.

Related Articles

Previous Articles