A federal judge expressed serious doubts about a $1.5 billion settlement between Anthropic and authors who claim their books were illegally used to train AI chatbots. Judge William Alsup worries about potential pitfalls, including the claims process and the involvement of publishers' groups, suggesting the settlement may not be approved. He scheduled a further hearing on September 25th to review the concerns, raising the possibility of a trial. The judge voiced concerns about the accuracy of the number of pirated books and the fairness of the payout process. The judge's skepticism has raised concerns that the settlement could fall apart.

Prepared by Jonathan Pierce and reviewed by editorial team.