JK Rowling could be in for a payout of thousands of pounds after Harry Potter’s publisher received a share of a billion-pound AI copyright settlement.
Anthropic, the company behind the Claude chatbot, will pay out $1.5bn (£1.1bn) to resolve claims that it used copyrighted books to train its AI models.
An estimated 482,000 works are covered by the settlement in total, including some 14,087 Bloomsbury-published titles.
Each book included in the copyright breach will receive $3,000 (£2,243) divided equally between publisher and author, the company confirmed on Wednesday.
Bloomsbury publishes Rowling’s books, as well as the best-selling novelists Sarah J Maas and Susanna Clarke.
The company expects to receive several rounds of payments, beginning later this year.
Rowling has refused to confirm whether she will benefit from the payout. The Harry Potter author would receive an estimated $1,500 (£1,120) per book if the series were included in the settlement.
The settlement, which took place in San Francisco, California, is the largest known copyright payout in US history.
The authors Andrea Bartz, Charles Graeber and Kirk Wallace Johnson first accused Anthropic of training its chatbot on pirated books in August 2024. The settlement was first proposed last September.
An estimated 482,000 works are covered by the settlement, and authors of about 91 per cent of the works involved have already made their claims, according to Anthropic.
Justin Nelson, the authors’ lawyer, said it was “the largest known copyright recovery in history”.
He said to AI Fray, an AI-regulation website, in October when the settlement was first proposed: “It is an amazing result. Being able to have accountability and to strike a blow on behalf of the creators is crucial.
‘Wake-up call to AI companies’
“The settlement itself sends a strong message that creators and copyright holders deserve compensation. It’s true in this case and true across the board, which is why it has been met with such praise from so many authors and creative community groups.”
He added: “The judgment on piracy really sent a strong message, and I hope it is a wake-up call to other AI companies that have used pirated datasets.”
Rowling has spoken out strongly for her intellectual property rights in the past.
She took out a lawsuit in 2008 against a fan’s plan to publish an A to Z of Harry Potter, calling the project “wholesale theft of 17 years of my hard work”.
The settlement comes weeks after Google renewed a campaign for ministers in the UK to relax AI copyright laws.
In June, the technology giant urged the Government to change the law to let tech start-up companies scrape the web’s trove of data to “train” their AI applications.
Recommended
Ministers consider letting AI dodge copyright rules
The training process involves AI companies harvesting publicly available data from the web, which are used to programme the model. This can include material protected by copyright.
Sir Paul McCartney and Sir Elton John are among the artists who have spoken out against loosening copyright protections for creative work in the UK.