A federal judge has raised significant concerns regarding a proposed $1.5 billion settlement between the artificial intelligence company Anthropic and a group of authors who allege that nearly 500,000 books were illegally pirated to train AI chatbots. U.S. District Judge William Alsup has indicated that the case could still proceed to trial, as he expressed doubts about the settlement's viability during a hearing held in San Francisco.
During the nearly hour-long hearing, Judge Alsup did not hold back his criticism of the settlement agreement, which was announced just days earlier by Anthropic and the attorneys representing the class-action lawsuit. The judge described the settlement as fraught with potential issues and scheduled a follow-up hearing for September 25 to determine if his concerns can be adequately addressed. “We’ll see if I can hold my nose and approve it,” Alsup remarked, indicating his reluctance to endorse the plan without further assurances.
In June, Judge Alsup issued a mixed ruling on the case, stating that while training AI chatbots on copyrighted material is not inherently illegal, Anthropic had unlawfully acquired millions of books from pirate websites to enhance its Claude chatbot. The proposed settlement aims to resolve the pirating claims and avoid a trial that was set to begin in December. Under the terms of the agreement, authors and publishers would receive approximately $3,000 for each book included in the settlement.
One of Alsup’s primary concerns revolved around the claims process and ensuring that all eligible authors are informed about their rights, so they do not end up “getting the shaft.” He has mandated a September 15 deadline for a definitive list of the books that were pirated, referred to as a “drop-dead list.” This list is crucial for preventing future lawsuits that could arise if the number of affected works increases unexpectedly.
Judge Alsup also expressed skepticism regarding the roles of two major organizations involved in the case: the Authors Guild and the Association of American Publishers. He raised concerns about these groups potentially pressuring authors into accepting the settlement without a full understanding of its implications. Notably, both Mary Rasenberger, CEO of the Authors Guild, and Maria Pallante, CEO of the Association of American Publishers, were present at the hearing but did not speak.
In the courtroom, the three authors who initiated the lawsuit—thriller novelist Andrea Bartz and nonfiction writers Charles Graeber and Kirk Wallace Johnson—sat in the front row but did not address the judge. Prior to the hearing, Johnson, known for his book “The Feather Thief,” described the settlement as the “beginning of a fight on behalf of humans that don’t believe we have to sacrifice everything on the altar of AI.”
During the proceedings, Justin Nelson, the attorney representing the authors, assured Judge Alsup that the distribution of the settlement funds would be fair. He pointed to the extensive media coverage of the case, which has included prominent stories in major newspapers, as evidence that the settlement process is well-publicized. “This is not an under-the-radar warranty case,” Nelson emphasized.
Despite the reassurances, Judge Alsup made it clear that he remains cautious about the settlement. He warned that if his concerns are not adequately addressed, he may opt to allow the case to proceed to trial. “I have an uneasy feeling about all the hangers on in the shadows,” he stated, underscoring the complexities surrounding the settlement and its implications for both the authors and the AI industry.