Anthropic pays 1.5 billion for pirated books

The historic $1.5 billion settlement in the copyright lawsuit between authors and the AI company Anthropic. The litigation centered on allegations that the developer utilized pirated datasets containing millions of books to train its Claude language model. While a judge previously suggested that AI training could qualify as fair use, he ruled that the use of illegally obtained materials did not deserve such protection. The resulting agreement provides approximately $3,000 per work to participating writers, though many authors and critics argue the amount is insufficient compared to the company’s massive valuation. Key details of the final approval include reduced legal fees, a 50/50 payment split between authors and publishers, and a requirement for Anthropic to delete the infringing data. Experts note that this case sets a significant legal precedent for how intellectual property is valued in the rapidly evolving artificial intelligence industry.