Judge William Alsup has rejected the record-breaking $1.5 billion settlement Anthropic has agreed to for a piracy lawsuit filed by writers. According to Bloomberg Law, the federal judge is concerned that the class lawyers struck a deal that will be forced “down the throat of authors.” Alsup reportedly felt misled by the deal and said it was “nowhere close to complete.” In his order, he said he was “disappointed that counsel have left important questions to be answered in the future,” including the list of works involved in the case, the list of authors, the process of notifying members of the class and the claim form class members can use to get their part of the settlement.

If you’ll recall, the plaintiffs sued Anthropic over the company’s use of pirated copies of their works to train its large language models. Around 500,000 authors are involved in the lawsuit, and they’re expected to receive $3,000 per work. “This landmark settlement far surpasses any other known copyright recovery,” one of the lawyers representing the authors said in a statement. However, Alsup had an “uneasy feeling about hangers on with all [that] money on the table.” He explained that class members “get the shaft” in a lot of class actions once the monetary settlement has been established and lawyers stopped caring.

Alsup told the lawyers that they must give the class members “very good notice” about the settlement and design a claim form that gives them the choice to opt in or out. They also have to ensure that Anthropic cannot be sued for the same issue in the future. The judge gave the lawyers until September 15 to submit a final list of works involved in the lawsuit. He also wrote in his order that the works list, class members list and the claim form all have to be examined and approved by the court by October 10 before he grants the settlement his preliminary approval.

Source: Judge rejects Anthropic’s record-breaking $1.5 billion settlement for AI copyright lawsuit

Of course this was only a small part of the actual lawsuit, which sought to establish that copyright precluded AIs from reading books without permission. This was struck down by the judge. The idiocy of Anthropic in using pirated books to train their AI beggars belief, but that is what they were punished for.

The reason the copyright lawsuit was put up was so that the copyright holders (the publishers, not the actual writers of the books – although that is what these publishers are telling you) could win megabucks. Now that the settlement has gone for piracy, the publishers and lawyers still want the megabucks, without sharing it with the actual writers. The judge says no.