Court’s Ruling on Anthropic AI Case May Have Ramifications for Music Industry

Judge William Alsup of the US District Court for the Northern District of California has ruled in the case brought by book authors against artificial intelligence startup Anthropic. As per _Music Business Worldwide_, he found that Anthropic’s “use of copyrighted books to train AI, without explicit permission to do so” counts as ‘fair use.’”

Behind the decision:

• Key to the Court’s ruling was the issue of whether use of the works is “transformative” – ie. is the output significantly different from the copyright material used to train the model.

• Alsup deemed Anthropic’s use “spectacularly transformative”, in that the product produced by Anthropic’s chatbot, Claude, was suitably different to the copyrighted source material.

• Therefore it was deemed fair use.

• The authors who sued Anthropic – Andrea Bartz, Charles Graeber and Kirk Wallace Johnson – did so not because they alleged Claude copied their books, but because they hadn’t given explicit permission for their books to be used to train Claude.

• The Judge dismissed the claim, stating: “Authors’ complaint is no different than it would be if they complained that training schoolchildren to write well would result in an explosion of competing works.”

• “The Copyright Act,” he added, “seeks to advance original works of authorship, not to protect authors against competition.”

How this relates to the music industry:

• Anthropic is also being sued by music publishers Universal Music Group (UMG), Concord, and ABKCO.

• As per MBW, their lawsuit is focused on the ‘outputs’ of Claude and the fact that Anthropic’s chatbot can “‘generate identical or nearly identical copies of [our] lyrics, in clear violation of publishers’ copyrights.’”

• UMG, Sony Music Entertainment and Warner Music Group make a similar claim in their lawsuit against AI music-making platforms Suno and Udio, alleging that the songs generated by these AI engines are “soundalikes” of existing copyrighted recordings.

The case continues:

• Anthropic isn’t out of the water yet. In training its AI model it download seven million books from online libraries “known to provide access to pirated copies of books.”

• The Judge ruled it was not “fair use” to copy and store pirated books for its central library.

• Anthropic will stand trial over the issue in December, facing penalties of up to $150,000 per infringement.

What’s next:

• Citing law experts from Law360, MBW points out that in light of Anthropic facing legal issues around its use of pirated books, “other AI developers may be spurred to sign licensing deals with rightsholders instead of risking hefty payouts.”

• It points to the recent reports of Suno and Udio entering into licensing talks with the record companies suing them as evidence.