"Data Please! Can disclaimers or terms & conditions really stop generative AI from scraping copyright protected works?"

In October 2024, Penguin Random House announced a significant policy change regarding the use of its literary works in the training of AI models. This move marks a notable shift in the publishing industry’s approach to the intersection of copyright law and AI development, reflecting growing concerns over the potential misuse of copyrighted materials and the ethical implications of AI-generated content. Can disclaimers help protect works from being used for unauthorised AI training?

What happened?

Penguin Random House has changed the wording on its copyright pages to help protect authors’ intellectual property from being used to train artificial intelligence tools.  The new wording states: “No part of this book may be used or reproduced in any manner for the purpose of training artificial intelligence technologies or systems.”

What does a disclaimer achieve?

Copyright protections exist irrespective of a disclaimer, and a disclaimer alone offers no additional legal protections. For disclaimers to offer legal protection, you need to make sure users accept them.

While a standalone disclaimer might not help you protect your works, what happens if that disclaimer is contained within terms & conditions that are accepted by a user?

Breach of contract vs copyright infringement

In the current legal grey area where AI exists, it’s much easier to sue for breach of contract than it would be to sue for copyright infringement, particularly in an area that is still untested in Australia.

Consider this scenario: If a website’s terms and conditions specify that users cannot use its content for AI training, and an AI company disregards this clause, voila! They’ve breached the contract. Cue damages! That’s a relatively straightforward case—at least until the AI company argues that it was the AI acting independently, not a human at the helm, that caused the breach. But let’s leave the debate over AI liability for another day.

Ethical AI?

The policy also serves as a proactive measure to clarify the publisher's stance on copyright protection, and the importance of authorial rights, emphasising that any AI training initiatives using its content will require explicit licensing agreements.

Policies such as these raise critical questions about how AI companies will adapt their strategies to comply with possible contractual restrictions and seek to encourage AI companies to behave in a ethical manner that respects the rights of creators.

Many publishers, authors and artists don’t have access to the resources to pursue expensive copyright litigation against a behemoth such a OpenAI – so, updating your terms & conditions to protect your intellectual property can be an easier, and cheaper, way to discourage unauthorised use of content.  

Questions? Give us a call.