Penguin Random House bans AI training on its books

Penguin Random House (PRH), the Anglo-American publishing giant, has announced a radical policy prohibiting the use of its books for training generative AI models. A new copyright clause will appear in all new and reprinted books stating: 'No part of this book may be used or reproduced in any manner for the purpose of training artificial intelligence technologies or systems.' The clause also explicitly reserves PRH's works from the 'text and data mining exception' under EU law. This move follows a statement from August 2024 in which PRH declared its intent to 'vigorously defend the intellectual property belonging to our authors and artists.' While PRH uses AI internally for sales forecasting and inventory management, it rejects creative AI applications and champions 'human creativity' and 'responsible innovation.' The legal enforceability of the clause remains uncertain, as international copyright laws on AI training are still evolving. The policy is seen as a strong statement and a 'nudge' toward new legislation, which may already lag behind technological advances. Other publishers have taken different stances: Wiley, Oxford University Press, and Taylor & Francis have all signed AI training agreements. The announcement comes amid growing backlash from artists and musicians—including Thom Yorke and Julianne Moore—who recently signed an open letter with 13,000 signatures against AI copying creative works.