Penguin Random House, the world's largest commercial publisher, plans to add language to the copyright pages of its books prohibiting the use of these books to train AI.
As The Bookseller reports, new books from publishers and reprints of older titles include: “No part of this book may be used or reproduced in any way for the purpose of training artificial intelligence techniques or systems.” It is said that the description will be included.
Although the use of copyrighted material to train AI models is currently being challenged in multiple lawsuits, Penguin Random House is the first to update its copyright page to reflect these new concerns. It seems like a major publisher.
This update itself does not necessarily change the legal status of the text (i.e. it does not mean that it was previously completely okay to use the book for AI training). Nor does it mean that Penguin Random House is completely opposed to the use of AI in book publishing.
In August, the publisher outlined its initial approach to generative AI, saying it would “strongly protect intellectual property belonging to authors and artists,” while also encouraging “the selective and responsible use of generative AI tools.” They also promised to help advance our goals.