Anthropic to Appeal Class Action Ruling in Copyright Lawsuit Over AI Training Practices

Anthropic PBC has announced its intention to appeal a recent decision by U.S. District Judge William Alsup, who certified a class action lawsuit against the company. The lawsuit, initiated by authors Andrea Bartz, Charles Graeber, and Kirk Wallace Johnson, alleges that Anthropic infringed upon their copyrights by using pirated copies of their books to train its AI language model, Claude.

Judge Alsup’s certification of the class action underscores the potential for significant liability for Anthropic. He noted that the company’s “Napster-style downloading of millions of works” from pirate libraries could be proven on a class-wide basis. The certified class includes legal or beneficial owners of copyrighted books with ISBN or ASIN numbers that Anthropic downloaded from pirate websites such as LibGen or PiLiMi. The trial is scheduled to commence on December 1, 2025.

In a prior ruling, Judge Alsup determined that Anthropic’s use of legally purchased books for training its AI model constituted fair use under the Copyright Act. However, he distinguished this from the company’s acquisition of other copyrighted works through downloading pirated materials, stating that such actions were not excused by their ultimate purpose of AI training.

Anthropic’s decision to appeal the class certification reflects the company’s ongoing legal strategy to challenge the allegations and mitigate potential damages. The outcome of this appeal could have significant implications for the broader AI industry, particularly concerning the use of copyrighted materials in training AI models.

As the case progresses, it will be closely watched by legal professionals and industry stakeholders, given its potential to set precedents in the intersection of copyright law and artificial intelligence development.