Anthropic payout piles pressure on UK ministers in AI copyright row
The publishing industry has urged ministers to force AI companies to license copyrighted content after Anthropic agreed a record $1.5bn (£1.1bn) settlement over its use of pirated books to train AI models.
Dan Conway, chief executive of the Publishers Association, told City PM the agreement should serve as a warning to AI developers operating in Britain, arguing the government must now introduce stronger rules to make sure tech firms pay creators for the material used to train their models.
“This is a significant moment in the ongoing global debate around copyright and AI,” Conway said.
“A major AI company has publicly agreed to compensate publishers and authors for using pirated material in its training data. This is a win for rightsholders with copyright registered in the US and, we hope, a signal to AI developers worldwide that they must obtain permission and license works for AI training.”
On Wednesday, a US federal court approved a $1.5bn class action settlement with Anthropic over its storage of millions of pirated books, ending one of the first major AI copyright disputes to result in compensation for authors and publishers.
14,000 books scraped
Bloomsbury Publishing confirmed on Tuesday that it is among the beneficiaries of the settlement, with more than 14,000 of its titles included.
The publisher expects to receive payments beginning later this year, with the settlement valuing eligible works at roughly $3,000 per title, before legal fees, to be split equally between authors and publishers.
The agreement follows a landmark ruling last year that Anthropic’s use of copyrighted books to train its AI models constituted fair use under US law.
However, the court separately found the company’s downloading and retention of millions of pirated books fell outside that protection, prompting Anthropic to settle rather than face a trial that could have exposed it to substantially larger damages.
“We reached this settlement in 2025, after the court’s landmark ruling that training AI on books is fair use under copyright law – which remains the law today,” said Aparna Sridhar, Anthropic’s deputy general counsel.
“We are pleased that more than 91 per cent of authors and publishers covered by the settlement have claimed their share of the payment, and we’re looking forward to bringing this matter to a close.”
Pressure grows on UK ministers
Conway said the case should not be interpreted as settling the wider legal argument over AI copyright training.
“The settlement relates only to the past use of pirated book content and does not establish that AI developers must obtain permission and pay for the use of all copyright-protected works in AI training,” he said.
He added that further litigation was needed to establish whether AI firms should need licences before using such material.
After abandoning proposals that would have allowed AI companies to train models on copyrighted works unless creators explicitly opted out, the government has instead committed to supporting licensing markets and exploring mandatory measures for AI developers.
Back in May, Baroness Keely, chair of the House of Lords Communications and Digital Committee, welcomed the government’s decision to drop its preferred opt-out model but warned that voluntary transparency was not enough.
“Only a mandatory framework will create the level playing field needed to foster responsible training data practices,” she said.
Jonathan Quinn, a partner specialising in AI litigation, said courts were only beginning to answer fundamental legal questions around generative AI.
“Artificial Intelligence is raising novel legal issues, many of which are now before the courts, but for most of which we do not yet have answers,” he said. “We are beginning to see courts provide some answers… For now, there are more questions than answers.”
The settlement comes as News Corp filed a lawsuit against AI search company Brave, alleging it copied and monetised the publisher’s journalism without permission, while separate copyright claims against OpenAI, Midjourney and AI music firm Suno continue to progress through the US courts.