Judge Rejects Anthropic $1.5B Copyright Settlement in AI Lawsuit

A federal judge, William Alsup, has denied preliminary approval of the $1.5 billion copyright settlement proposed by AI company Anthropic and book authors involved in a class action lawsuit, citing concerns about the fairness and transparency of the deal. Alsup expressed unease over the settlement negotiations occurring behind closed doors without adequate input from the authors, warning that class members often receive poor outcomes once deals are struck and lawyers lose interest. Key issues requiring clarification include the list of works covered, notification processes for class members, and the claims filing procedure, all of which must be resolved before the settlement can move forward. The lawsuit originated from allegations that Anthropic used millions of pirated books to train its AI models, a claim reflecting broader legal challenges faced by tech firms in the generative AI space. This case marks a significant moment in AI copyright litigation, potentially shaping future market-based licensing approaches for training data. Judge Alsup has set deadlines for the parties to provide the necessary details and documentation to ensure fair treatment of the authors before any approval is granted.