Authors involved in a class-action lawsuit over Anthropic’s AI training data say a “historic” settlement is imminent, marking a major turn in a case that could affect potentially millions of writers. According to Ars Technica, US District Judge William Alsup confirmed that Anthropic and the authors “believe they have a settlement in principle” and plan to seek preliminary approval by September 5.
Judge confirms settlement-in-principle after massive class certification
The announcement follows Alsup’s certification of what AI industry advocates criticized as the largest copyright class action to date. Although the lawsuit was initiated by authors Andrea Bartz, Kirk Wallace Johnson, and Charles Graeber, the court allowed up to 7 million potential claimants to join, reflecting the number of books that Anthropic may have downloaded to train its models. Industry advocates warned that full participation could “financially ruin” the AI industry.
Specific settlement terms remain under seal, but court filings indicate they are binding. Justin A. Nelson, counsel for the authors, told Ars that details will be shared soon and said the suing authors are claiming a win on behalf of possibly millions of class members. “This historic settlement will benefit all class members,” Nelson said, adding that further information would be announced in the coming weeks.
Legal stakes, timing, and what’s still unresolved
Ars reported that Anthropic previously argued the lawsuit could doom the company, which was founded in 2021 by former OpenAI employees. Facing trial in four months, the company confronted “hundreds of billions of dollars in potential damages liability,” and Wired estimated potential exposure could exceed “more than $1 trillion in damages.”
Fair use ruling and strategic shift
Alsup had earlier ruled that Anthropic’s training on authors’ works was “fair use,” suggesting the settlement will not resolve broader copyright questions still circulating in the AI sector. Advocates had warned that certifying such a large class could pressure AI companies to settle or even fold when facing substantial damages risks, regardless of the merits.
Wired quoted Edward Lee, a law professor at Santa Clara University, who called the development “a stunning turn of events, given how Anthropic was fighting tooth and nail in two courts in this case. And the company recently hired a new trial team.” Lee said Anthropic “had few defenses at trial, given how Judge Alsup ruled,” leaving the company exposed to statutory damages in “doomsday” amounts. Ars noted it could not immediately reach Anthropic for comment.
