Courts on AI / Substantive AI law courts cite
Bartz v. Anthropic PBC
U.S. District Court for the Northern District of California · Federal district courts · CA · · No. 3:24-cv-05417-WHA · order
The Northern District of California rules on Anthropic’s book copying for Claude and related language models. Judge Alsup distinguishes training from maintaining a central library, finding transformative fair use for training while allowing claims about pirated library retention to proceed.
- Disposition
- The court ruled that training uses were fair use but preserved claims over Anthropic’s central library.
What the court wrote
Verbatim from the court's document (public domain). Ellipses mark omitted text; nothing is paraphrased.
An artificial intelligence firm downloaded for free millions of copyrighted books in 17 digital form from pirate sites on the internet. The firm also purchased copyrighted books 18 (some overlapping with those acquired from the pirate sites), tore off the bindings, scanned 19 every page, and stored them in digitized, searchable files. All the foregoing was done to amass 20 a central library of “all the books in the world” to retain “forever.” From this central library, 21 the AI firm selected various sets and subsets of digitized books to train various large language 22 models under development to power its AI services. Some of these books were written by 23 plaintiff authors, who now sue for copyright infringement. On summary judgment, the issue is 24 the extent to which any of the uses of the works in question qualify as “fair uses” under 25 Section 107 of the Copyright Act. 26 STATEMENT 27 Defendant Anthropic PBC is an AI software firm founded by former OpenAI employees 28 in January 2021. Its core offering is an AI software service called Claude. When a user Case 3:24-cv-05417-WHA Document 231 Filed 06/23/25 Page 2 of 32 1 prompts Claude with text, Claude quickly responds with text — mimicking human reading and 2 writing. Claude can do so because Anthropic trained Claude — or rather trained large 3 language models or LLMs underlying various versions of Claude — using books and other 4 texts selected from a central library Anthropic had assembled.