MUMBAI: The legal chapter may be closing, but the AI copyright story is far from its final draft. A US federal court has given final approval to Anthropic’s $1.5 billion copyright settlement, bringing one of the biggest legal battles over generative AI training a step closer to resolution and marking what is widely regarded as the largest copyright settlement in US legal history.
The ruling by US District Judge Araceli Martínez-Olguín in San Francisco clears the way for compensation to thousands of authors who accused the AI company of using their books without permission to train its Claude chatbot. The court rejected objections from authors who argued the settlement undervalued their claims or disproportionately rewarded the lawyers representing the class.
The lawsuit, filed in 2024, alleged that Anthropic relied on pirated copies of books to train its large language models, making it one of the earliest and most closely watched copyright disputes involving generative AI.
The settlement follows a landmark decision last year by then-presiding Judge William Alsup, who ruled that using copyrighted books to train AI models qualified as fair use under US copyright law. However, he also found that Anthropic had infringed copyright by maintaining a central digital library containing more than seven million pirated books, a collection that was not necessarily limited to AI training purposes.
That ruling left Anthropic facing a separate trial over potential damages linked to the storage of those works, with possible financial exposure estimated to run into hundreds of billions of dollars before the parties reached a settlement.
Following the court’s approval, Anthropic said the agreement reflected the legal framework established by the earlier fair use ruling and revealed that more than 91 per cent of eligible authors and publishers had already claimed their share of the settlement funds.
The court also approved legal fees exceeding $101 million, although that was lower than the $187.5 million sought by the plaintiffs’ legal team.
Judge Martínez-Olguín dismissed arguments that the settlement amount was inadequate, saying the objections failed to sufficiently account for the legal uncertainty and risks that would have accompanied a full trial.
The agreement does not end Anthropic’s legal challenges entirely. A number of authors and publishers opted out of the class settlement and continue to pursue separate copyright claims, ensuring that the broader legal debate over how AI companies use copyrighted material to train their models remains far from settled.