Anthropic, the AI company behind the chatbot Claude, has agreed to the largest copyright infringement settlement in U.S. history, paying $1.5 billion to settle a lawsuit involving nearly 500,000 writers.
Anthropic, the AI company behind the chatbot Claude, has agreed to the largest copyright infringement settlement in U.S. history, paying $1.5 billion to settle a lawsuit involving nearly 500,000 writers.
As reported by the New York Times, the court found Anthropic guilty of illegally downloading books and using them to train its artificial intelligence model. The total amount of payments in this case is the largest in US history among all copyright infringement cases, although the payment to each affected author is only $3,000.
In a lawsuit filed in August 2024, Anthropic was accused of profiting from pirated copyrighted books. The court found that when Anthropic legally acquired copyrighted books, the law allowed the company to train artificial intelligence technology using those books because it transformed them into something new.
However, the judge ruled that Anthropic also illegally obtained millions of books through online libraries such as Library Genesis and Pirate Library Mirror, which were used by many tech companies to train artificial intelligence technologies.
«Anthropic’s LLMs Claude jeopardize the ability of authors to earn a living because LLMs allow anyone to automatically and for free (or very cheaply) generate texts that authors would otherwise be paid to create and sell. Anthropic’s LLMs, which erode the commercial market for plaintiffs’ and class’s works, were created without paying the authors a single cent. Anthropic’s enormous success is a direct result of copyright infringement,» the authors’ class action lawsuit states.
The judge ruled that when Anthropic downloaded the illegal libraries, its executives knew they contained pirated books. The company could have purchased the books from many sellers, the judge noted, but instead decided it was better to «steal» them to avoid «legal/practical/business routine,» according to Anthropic CEO Dario Amodei.
After the judge ruled in favor of the writers, the two sides agreed to a settlement. Anthropic said it did not use pirated works to create its AI technology, but agreed to remove pirated books it had downloaded and stored.
Experts say the deal could force more tech companies to pay copyright owners through court rulings and settlements. AI company Perplexity, for example, is already offering publishers the chance to receive a share of the revenue their articles generate when they are found by Comet’s AI search engine.
Recall that in early September , Anthropic received $13 billion in new funding. This raised its valuation to $183 billion and made it one of the most valuable artificial intelligence companies in the world.
