Anthropic is seeking to shut down a copyright case brought by Universal Music Group and other publishers over the use of song lyrics in training its Claude chatbot, arguing that the process is legally transformative and therefore protected by fair use. In a filing on Monday, the company said the publishers could not seriously dispute that teaching a model with lyrics and other text is different from reproducing the works themselves.

The dispute turns on one of the most contested questions in artificial intelligence law: whether training on copyrighted material counts as fair use. Anthropic says Claude absorbs lyrics alongside trillions of other words in order to recognise patterns in language, support coding, research and document drafting, and generate outputs that are largely unrelated to music. Its lawyers say that makes the model a general-purpose tool rather than a substitute for songs or lyrics.

Anthropic also argues the publishers have not shown meaningful market harm, another factor a court may weigh in assessing fair use. The company pointed to remarks by Universal Music Group chief digital officer Michael Nash, who told investors last month that AI’s effect on the business would be "overwhelmingly net positive", a comment Anthropic cites as undercutting the labels’ claim of damage.

The publishers, however, rejected the filing. In a statement to Billboard on Tuesday, a representative said there was "no excuse" for what they described as blatant infringement, adding that Anthropic was wrong on both the facts and the law and that the companies will respond in detail in their opposition brief. The case, first filed in 2023, sits within a wider wave of litigation over AI training, including separate music-industry suits against Suno and Udio that remain unresolved despite some licensing deals struck late last year.

Source Reference Map

Inspired by headline at: [1]

Sources by paragraph:

Source: Noah Wire Services