Anthropic settles AI copyright lawsuit for $1.5 billion, but authors describe outcome as a mixed result

A federal judge in San Francisco approved a $1.5 billion settlement in July resolving a class action lawsuit brought by authors against AI company Anthropic roughly two years after the case was filed. The suit alleged that Anthropic used the contents of millions of digitized copyrighted books to train the large language models behind its Claude chatbot without the consent of or payment to the authors.

Lead plaintiff Charles Graeber, a New York Times bestselling author whose book The Good Nurse was adapted into a Netflix film in 2022, said he expects to receive roughly $3,100 per book used in training Claude. He described a degree of pride in the outcome — characterizing it as a "ragtag bunch of authors" holding together as a class against a large technology company — but said the personal costs were significant. "I'm much poorer for this settlement, ironically," he told NPR, citing years of travel, time, and work passed up during the litigation.

The settlement's total figure is divided among hundreds of thousands of claimants, shared with publishers, and reduced further by more than $100 million in attorneys' fees, according to NPR's reporting. Co-plaintiff Andrea Bartz said that while the settlement establishes that unauthorized use carries a price, the court's narrow ruling stopped short of declaring AI training on copyrighted works illegal — a distinction she said could favor AI companies in future litigation.