Anthropic Judge Rejects $1.5B AI Copyright Settlement (news.bloomberglaw.com)

A federal judge has halted approval of Anthropic PBC’s proposed $1.5 billion copyright settlement, citing concerns that class attorneys may be pushing the deal on authors without sufficient transparency or input. Judge William Alsup of the Northern District of California expressed unease over an incomplete settlement framework, particularly criticizing the lack of clarity on how potential claimants will be notified and how claims will be processed. He stressed the importance of robust notice procedures to protect both class members’ rights and Anthropic from future litigation, instructing parties to submit a comprehensive list of covered works and a clear opt-in claim form by September 15. This decision is significant for the AI and legal communities as the settlement represents one of the largest copyright resolutions involving AI training data to date, setting a potential precedent for how AI companies handle copyright disputes stemming from using vast copyrighted materials like books. The complexities raised by Judge Alsup—including multi-author and publisher ownership issues—highlight the challenges in structuring fair settlements in AI copyright cases. The judge emphasized that any works with disputed ownership should be resolved in state courts, underscoring the intricacies of intellectual property law in the AI domain. The ruling forces involved parties to refine their approach and may delay a final resolution, reflecting broader tensions around AI’s use of copyrighted data and the legal frameworks needed to manage it. Stakeholders like the Association of American Publishers criticized the court’s skepticism over the claims process, revealing diverging views on balancing author compensation with workable legal procedures. As AI systems continue to rely on expansive content licensing, this case will likely influence how future copyright settlements in AI training datasets are negotiated and approved.