Background on Anthropic and the Relevant Authors
Anthropic, a prominent artificial intelligence (AI) company, has agreed to pay at least $1.5 billion to settle a class-action lawsuit in the United States for using 500,000 books without permission from their creators to train its AI model Claude. The authors involved in the lawsuit are Andrea Bartz, Charles Graeber, and Kirk Wallace Johnson. Their accusations claim that Anthropic illegally copied their books to train Claude, an AI chatbot competing with ChatGPT.
Legal Background and Court Decisions
In June, Anthropic secured a partial victory when U.S. federal judge William Alsup of San Francisco ruled that training AI models with books, whether purchased or pirated, transformed the works to such an extent that it constituted a “fair use” under U.S. copyright law. However, the judge determined that Anthropic’s practice of downloading millions of pirated books to build a digital library did indeed infringe on copyright laws.
Details of the Settlement
The settlement covers approximately 500,000 books, equating to around $3,000 per work—four times the minimum legal damages for copyright infringement under U.S. law. Anthropic will destroy the original pirated files and any derivative copies, but it retains rights to legally purchased and scanned books.
Impact on the AI Industry
Mary Rasenberger, CEO of the Authors Guild, stated in a press release: “This settlement sends a strong message to the AI industry: there are serious consequences when authors’ works are pirated to train AI.” The settlement still requires approval from Judge Alsup, with a hearing scheduled for Monday at the U.S. District Court in San Francisco.
Key Questions and Answers
- Who is Anthropic? Anthropic is a well-known artificial intelligence company that specializes in developing advanced AI models, including Claude, an AI chatbot competing with ChatGPT.
- Why is this settlement significant? This settlement is the first of its kind in the era of AI, surpassing any previous copyright recovery. It highlights the serious consequences of using pirated works to train AI models.
- What is the total amount Anthropic will pay? Anthropic has agreed to pay at least $1.5 billion as part of the settlement.
- What specific actions led to this lawsuit? Anthropic used 500,000 books without permission from their creators to train its AI model Claude.
- What will Anthropic do with the pirated files? Anthropic will destroy the original pirated files and any derivative copies, while retaining rights to legally purchased and scanned books.
- What is the next step in this legal process? The settlement still requires approval from Judge Alsup, with a hearing scheduled for Monday at the U.S. District Court in San Francisco.
