The lawsuit, filed by authors including Andrea Bartz, accuses Anthropic—backed by Amazon and Alphabet—of using pirated copies of books without consent
A federal judge ruled Monday that Anthropic’s use of copyrighted books to train its AI model was “fair use” under US law—but also found the company infringed authors’ rights by storing over 7 million pirated copies.
Judge William Alsup said training Claude AI was “transformative” and supported innovation.
However, Alsup also found that Anthropic infringed copyright laws by storing at least 5 million and 2 million unauthorized copies of books from LibGen and PiLiMi, which he said was not fair.
The lawsuit, filed by authors including Andrea Bartz, accuses Anthropic—backed by Amazon and Alphabet—of using pirated content without consent. The ruling could shape the future of AI and the boundaries of copyright.
Similar cases are underway against OpenAI, Meta, and Microsoft. Anthropic claims its practices are legal and promote technological progress.