Thousands of authors speak out against using their works for AI training

Around 10,000 writers including Kazuo Ishiguro, Philippa Gregory and Richard Osman participated in publishing an ‘empty’ book at the London Book Fair to protest against using their works for AI training.

The only content of Don’t Steal This Book is a list of authors who have joined the protest. Copies are being given out for free to visitors at the fair a week before March 18, 2026 when the British government will decide on changes to copyright law. The book’s back cover says: “The UK government must not legalise book theft to benefit AI companies.”

The campaign was organized by composer and campaigner for protecting artists’ copyright Ed Newton-Rex who said that the AI industry was “built on stolen work … taken without permission or payment.”

RELATED: Librarians face growing requests for AI-invented books

“This is not a victimless crime – generative AI competes with the people whose work it is trained on, robbing them of their livelihoods. The government must protect the UK’s creatives, and refuse to legalise the theft of creative work by AI companies,” added Newton-Rex.

The government’s main proposal is to permit AI companies to use copyrighted works without obtaining permission from the copyright holders, unless the copyright holder has stated their refusal.

Three other options are also being considered: 

• Maintaining the status quo, introducing mandatory licensing;
• Permitting the use of works without any possibility of refusal by authors;
• The government may also abolish copyright for materials for the purpose of “commercial research” — representatives of the creative industries fear that this will give AI companies access to their works without permission or payment.

“The government wants a copyright regime that values and protects human creativity, can be trusted, and unlocks innovation. We will continue to engage closely with the creative sector on this issue, and we will meet our commitment to update parliament by March 18th,” a government spokesperson said.

At the same time, the non-profit organization Publishers’ Licensing Services (PLS) launched a collective licensing initiative at the fair. Via an online content store, AI companies will be able to legally access publishing materials under certain conditions. The initiative welcomes publishers of different sizes, such as book, academic, and media publishers.

Last year Anthropic agreed to pay $1.5bn to settle a class-action lawsuit filed by writers who claimed that the startup had used pirated copies of their books to train chatbot Claude. At the same time, the writers sued Apple for similar use of their works.

In 2025, British authors protested outside Meta’s London office over the use of more than 7.5 million books from the LibGen “shadow library” to train AI models. In February, European publishers filed an antitrust complaint with the European Commission against Google for using their content in AI services without consent or payment.

RELATED: The London Book Fair 2025: ‘Romantasy,’ AI, audiobooks, ‘new reader’ and reading for pleasure

Main image: Jill Mead/The Guardian