A US federal judge in San Francisco has ruled that Anthropic’s use of copyrighted books to train its AI system, Claude, constitutes “fair use” under US copyright law. Judge William Alsup sided with the tech company, dismissing claims from authors like Andrea Bartz, Charles Graeber, and Kirk Wallace Johnson, who argued their works were used without permission.
This decision marks a significant victory for the AI industry, addressing a key concern regarding the legality of using copyrighted material for training large language models. However, the ruling also highlighted a potential problem for Anthropic. Judge Alsup noted the company’s “central library” containing over seven million pirated books. The judge did not make a ruling based on this pirated work.
