A federal judge has sided with Anthropic in an AI copyright case, ruling that training — and only training — its AI models on legally purchased books without authors’ permission is fair use. It’s a first-of-its-kind ruling in favor of the AI industry, but it’s importantly limited specifically to physical books Anthropic purchased and digitized.
Judge William Alsup of the Northern District of California also says in his decision that the company must face a separate trial for pirating “millions” of books from the internet. The decision also does not address whether the outputs of an AI model infringe copyrights, which is at issue in other related cases.
The lawsuit was filed by writers Andrea Bartz, Charles Graeber, and Kirk Wallace Johnson, who sued Anthropic last year over claims the company trained its family of Claude AI models on pirated material. It’s a pivotal decision that could affect how judges respond to AI copyright cases going forward.
The ruling also addresses Anthropic’s move to purchase print copies of books, rip off their bindings, cut the pages, and scan them into a centralized digital library used to train its AI models. The judge ruled that digitizing a legally purchased physical book was fair use, and that using those digital copies to train an LLM was sufficiently transformative to also be fair use.
“Authors’ complaint is no different than it would be if they complained that training schoolchildren to write well would result in an explosion of competing works,” Judge Alsup writes, adding that the Copyright Act “seeks to advance original works of authorship, not to protect authors against competition.”
Despite these wins for Anthropic, Judge Alsup writes that Anthropic’s decision to store millions of pirated book copies in the company’s central library — even if some weren’t used for training — isn’t considered fair use. “This order doubts that any accused infringer could ever meet its burden of explaining why downloading source copies from pirate sites that it could have purchased or otherwise accessed lawfully was itself reasonably necessary to any subsequent fair use,” Alsup writes (emphasis his).
Judge Alsup says the court will hold a separate trial on the pirated content used by Anthropic, which will determine the resulting damages.
