A U.S. federal judge has approved a $1.5 billion settlement between artificial intelligence company Anthropic and a group of authors who accused the company of using pirated copies of their books to train its AI chatbot, Claude, marking what plaintiffs described as the largest copyright recovery in U.S. history.
Judge Araceli Martínez-Olguín of the U.S. District Court in San Francisco granted final approval to the class-action settlement on Monday, July 20. The agreement covers more than 482,000 books, with authors and publishers set to receive approximately $3,000 per book. According to Anthropic, more than 91% of eligible rights holders have already submitted claims for compensation.
The lawsuit was filed in 2024 by authors including Andrea Bartz, Charles Graeber and Kirk Wallace Johnson, who alleged that Anthropic downloaded millions of copyrighted books from piracy websites to build a central digital library used in developing its Claude AI models.
The settlement follows a key ruling issued last year by U.S. District Judge William Alsup, who found that using copyrighted books to train AI models could qualify as fair use under U.S. copyright law. However, he also ruled that Anthropic’s acquisition and storage of books obtained from piracy websites did not qualify for fair-use protection, leaving the company exposed to potential copyright damages.
In a statement cited by Reuters, Anthropic said it was pleased that the vast majority of eligible authors and publishers had claimed their share of the settlement and looked forward to bringing the case to a close.
The settlement resolves one of the first major class-action copyright disputes involving generative AI training data. However, it does not end Anthropic’s legal challenges, as several authors and publishers who opted out of the agreement continue to pursue separate copyright lawsuits against the company.
