Anthropic Reaches Landmark $1.5 Billion Settlement in AI Training Copyright Lawsuit

A U.S. federal judge in San Francisco has given final approval to a historic $1.5 billion settlement between artificial intelligence company Anthropic and a group of authors. The lawsuit accused Anthropic of improperly using authors' books to train its AI chatbot, Claude. This settlement, approved by U.S. District Judge Araceli Martinez-Olguin, is reportedly the largest ever in a U.S. copyright case. The action is one of many filed by copyright holders, including writers and media outlets, against tech companies for using protected works in training large language models. It marks the first major case of its kind in the U.S. to reach a settlement. The authors initially sued Anthropic in 2024, alleging that the company, which is backed by Amazon and Alphabet, used pirated versions of their books without authorization to teach Claude how to respond to user prompts. While a previous judge, William Alsup, had initially ruled that using the works for training might fall under U.S. "fair use" law, he also found that Anthropic violated copyright by storing over 7 million pirated books in a central library, regardless of whether they were used for AI training. The trial to determine damages, which could have reached hundreds of billions of dollars, was scheduled to begin in December. According to one of the authors' lawyers, over 92% of the approximately 480,000 works covered by the settlement have had claims filed. However, the settlement has faced criticism, with some authors deeming the amount insufficient, arguing that legal counsel will receive an outsized portion, or that certain copyright holders were unfairly excluded from negotiations. Additionally, some writers and publishers have opted out of the settlement and pursued their own separate legal actions against Anthropic, which remain ongoing.