On September 5, 2025, a landmark settlement was announced in the case Bartz et al. v. Anthropic PBC, No. 3:24-cv-05417-WHA (N.D. Cal.), marking the largest publicly reported copyright settlement in U.S. history and the most significant resolution to date of a dispute involving the use of copyrighted works to train artificial intelligence models.
The litigation was initiated by authors Andrea Bartz, Kirk Wallace Johnson, and Charles Graeber, who alleged that Anthropic PBC had unlawfully downloaded and used copyrighted books from the notorious pirate repositories Library Genesis (“LibGen”) and Pirate Library Mirror (“PiLiMi”). These “shadow libraries,” repeatedly targeted by law enforcement and courts, are known for providing unauthorized copies of hundreds of thousands of books.
Following the certification of the “LibGen & PiLiMi Pirated Books Class” by U.S. District Judge William Alsup, it was revealed that Anthropic had accessed and utilized works from these illegal sources in the course of developing its AI systems. The settlement was reached after extensive negotiations involving authors, publishers, and their representatives.
The proposed settlement requires Anthropic to pay $1.5 billion plus interest into a settlement fund. Based on current estimates, this equates to approximately $3,000 per work for nearly 500,000 works included in the certified class. Should additional works be added, Anthropic will be required to pay $3,000 per additional work, with the possibility of higher recovery depending on the number of claims submitted.
Other critical aspects of the agreement include:
This resolution is widely regarded as a turning point in copyright litigation and AI regulation. Beyond the monetary compensation, the settlement sends a strong message to AI developers that reliance on pirated content as training data will not be tolerated and carries substantial financial consequences.
Maria Pallante, President and CEO of the Association of American Publishers, stated:
“I believe that the settlement as presented is beneficial to all class members and I am hopeful that the settlement will receive wide support from copyright owners. Beyond the monetary terms, the proposed settlement provides enormous value in sending the message that Artificial Intelligence companies cannot unlawfully acquire content from shadow libraries or other pirate sources as the building blocks for their models.”
Mary Rasenberger, CEO of the Authors’ Guild, similarly emphasized:
“An excellent result for authors, publishers, and rightsholders generally, sending a strong message to the AI industry that there are serious consequences when they pirate authors’ works to train their AI, robbing those least able to afford it.”
The settlement remains subject to court approval. A hearing on preliminary approval is scheduled for September 8, 2025, before Judge Alsup. Should preliminary approval be granted, notice will be provided to potential class members via the dedicated website AnthropicCopyrightSettlement.com. A final approval hearing could be scheduled in 2026.
Class counsel appointed by the court are Justin Nelson (Susman Godfrey LLP) and Rachel Geman (Lieff Cabraser Heimann & Bernstein LLP). The settlement website will provide a searchable list of covered works and detailed instructions for class members regarding their rights and options.
This historic agreement underscores the importance of compliance with copyright law in the age of artificial intelligence. It demonstrates that creators—including authors, musicians, artists, and journalists—can obtain significant remedies when their works are exploited without authorization.
For AI companies, the settlement highlights the urgent need to ensure that training datasets are lawfully acquired, contractually licensed, and fully compliant with intellectual property rights. The use of pirated materials from shadow libraries carries not only reputational risks but also potentially unprecedented financial liability.
Our firm closely monitors these developments and advises clients—both rightsholders and technology companies—on the full spectrum of intellectual property issues raised by artificial intelligence.
