A San Francisco federal judge approved a $1.5 billion settlement paid by Anthropic to authors who alleged the company misused their books. The class action lawsuit targeted the training data used for the Claude chatbot. This massive payout sets a costly precedent for how AI labs compensate creators for training data.