One of the most important copyright cases of the AI era, Bartz v. Anthropic, shows that authors’ rights cannot be ignored.
The case was brought forward in the summer of 2024 by authors Andrea Bartz, Charles Graeber, and Kirk Wallace Johnson, who alleged that Anthropic pirated millions of books from illegal shadow libraries such as LibGen and PiLiMi to train its AI tool Claude.
In June 2025, Judge William Alsup of the U.S. District Court for the Northern District of California ruled on summary judgment that using books without permission to train AI models could qualify as fair use, but only if the books were acquired legally. However, he denied the request from Anthropic for summary judgment on the piracy claims, finding that copying books from illegal sources was not fair use.
On July 17, 2025, Judge Alsup certified the piracy claims as a class action, including both authors and publishers whose works had been pirated, provided the books were registered with the U.S. Copyright Office and had ISBN or ASIN numbers.
Following mediation, the parties reached a proposed settlement, which was preliminarily approved on September 25, 2025. Anthropic agreed to pay $1.5 billion and to destroy the pirated datasets once litigation obligations end.
This case is a historic affirmation that authors’ work has real value and that copyright protections cannot be bypassed, even in the rapidly evolving AI landscape. It underscores the importance of defending creative rights and ensures that authors are recognised and compensated for their contributions even in this new digital age.
Read the range of responses and official statements from IAF U.S. members regarding this important news: