Hajra Naz
A U.S. federal judge has ruled that Anthropic’s use of books to train its AI model was legal under U.S. copyright law, according to a report by Reuters. The ruling marks a major win for the tech industry in the ongoing debate over how copyrighted content can be used to train artificial intelligence systems.
Judge William Alsup, based in San Francisco, said that Anthropic’s use of books by authors Andrea Bartz, Charles Graeber, and Kirk Wallace Johnson qualified as “fair use.” The company had used these books to train its Claude large language model, one of the leading AI systems developed by Anthropic.
However, the ruling was not entirely in Anthropic’s favor.
The judge also found that Anthropic’s copying and storing of over 7 million books in what he described as a “central library” did violate copyright laws. That action, he said, did not qualify as fair use. A trial will now take place in December to determine how much Anthropic must pay in damages. U.S. copyright law allows statutory damages of up to $150,000 per work in cases of willful infringement.
In a statement, Anthropic said it was pleased the court recognized that training its AI was “transformative” and aligned with the broader goals of copyright—to promote creativity and scientific progress.
The lawsuit was originally filed last year by the authors. They claimed Anthropic used unauthorized copies of their books without permission or payment to help train Claude. The case is supported by a group of writers and could turn into a class action.
It’s one of many similar lawsuits filed against AI companies like OpenAI, Microsoft, and Meta, all of which are facing legal scrutiny over how they use copyrighted material to train large language models.
The idea of fair use—which allows copyrighted content to be used without permission in certain cases—is now central to the defense strategies of these tech firms. This decision from Judge Alsup is the first of its kind to directly address fair use in the context of generative AI.
AI companies argue that their systems create new, transformative content from existing material and that requiring payment for all source content could stifle innovation and slow down progress in the fast-growing AI field.
This ruling is expected to influence how similar cases are handled going forward. It also highlights the legal gray areas that still surround AI, content ownership, and copyright in the digital age.
