Federal Judge Approves $1.5B Anthropic Copyright Settlement

This settlement establishes concrete liability and payment mechanisms for unauthorized use of copyrighted material in AI training, directly affecting Anthropic's operating costs ($1.5B payout over two years) and creating precedent expectations for ongoing litigation against Google, Meta, Midjourney, and OpenAI. The ruling that *obtaining* books from pirate sites violates copyright—while training on copyrighted text may be fair use—splits the legal exposure for AI companies: acquisition method now carries measurable liability even if training itself survives fair-use defense. Publishers and authors gain a defined compensation model; the broader AI industry faces 91% class participation signaling content-holder appetite for licensing vs. open litigation.