A federal judge in San Francisco approved a $1.5 billion settlement in July between hundreds of thousands of authors and AI company Anthropic, resolving a class action lawsuit filed approximately two years prior. The settlement came after a previous ruling found that Anthropic had used millions of digitized copyrighted books to train its Claude chatbot without author authorization or compensation.
Charles Graeber, one of three main plaintiffs and a New York Times bestselling author, expressed complicated emotions about the settlement despite receiving approximately $3,100 per book used. While he viewed the outcome as validation that piracy constitutes a crime, he noted the personal costs of pursuing the case over more than two years, including forgone career opportunities. The settlement amount must be divided among over 300,000 authors, with publishers also claiming shares and plaintiffs’ lawyers receiving more than $100 million of the total.
The ruling carries ambiguous implications for copyright protections in AI training. Although Anthropic paid the settlement, the court did not declare AI training on copyrighted works illegal under fair use doctrine. Similar rulings favored Meta in a separate author lawsuit, establishing precedent that fair use may apply when copyright holders cannot demonstrate market harm. Andrea Bartz, another named plaintiff, contested this interpretation, arguing that AI-generated text directly competes with human authors’ work.
Global enforcement challenges complicate potential protections for American authors. Chinese AI companies often employ “AI distillation,” using outputs from U.S. models rather than directly training on pirated books, complicating copyright enforcement efforts. Bartz proposed international copyright treaties to address cross-border concerns.
Licensing agreements between publishers and AI companies represent a potential alternative path forward, though adoption remains limited due to ongoing litigation. Industry advocates note that licensing enables content creators to control how their work appears in AI outputs while ensuring legal training practices.
Article Attribution | Read More at Article Source
Article summary produced by Claude AI