AI makers have faced a lot of criticism for borrowing from the work of others to train their models. Now the world’s largest publishing house is taking steps to prevent its authors from plagiarizing their works in the name of progress.
Penguin Random House Publishing has changed the copyright page on the front of their books to use any name as a resource for AI training. The text now reads: “No part of this book may be used or reproduced in any way for the purpose of teaching artificial intelligence technologies or systems.”
The new text also protects against data mining by noting the publisher “explicitly cautions [the titles] except for text and data mining.” This part of the modified text comes from the last time text and data mining exceptions and ownership.
Penguin Random House is the latest publishing company to dabble in AI models. Earlier this week Perplexity has issued a cease-and-desist letter to an artificial intelligence startup to detect the use of its articles and stories to help its AI model generate answers for users.
