Anthropic Settlement - FAQs for Cambridge authors

What is the Anthropic Settlement?

You may have read in the news that Anthropic, the AI company which created chatbot Claude, has agreed to pay $1.5 billion to settle a copyright infringement lawsuit instigated by authors in the US. Anthropic was being sued for using copyright-protected books without permission to train its AI model. 

Was my copyright work infringed by Anthropic? 

The court is currently reviewing the list of about 500,000 registered copyright books alleged to have been used to train Anthropic’s AI model. This list of works has recently been finalised by the Court and there are books on the list  which were published by or associated with Cambridge University Press. We will communicate with the authors of any relevant works in due course.

How is Cambridge responding? 

Cambridge is currently working closely with various industry bodies such as the Publishers Association and Association of University Presses to understand and act upon communications from Publisher’s Co-Ordination Counsel (PCOC). PCOC are our lawyers in this case who are advising us on how best to identify Cambridge University Press books which are involved and make any applicable claims for compensation. None of the works identified as part of the settlement have originated from Cambridge journals content.

Will Cambridge works that have been infringed be compensated as part of the settlement?

In addition to having an identifiable ISBN, any work to be included on the final list also needs to meet two other criteria in order to be eligible for compensation: firstly, to have been downloaded by Anthropic from the LibGen or PiLimi shadow libraries in the period from June 2021 to August 2022 and, secondly, to have been registered as a copyright work in the US. 

Was my work registered by Cambridge with the US Copyright Office?

In line with common practice with other international publishers, copyright works created by Cambridge authors are not registered as a matter of course in the US.

Copyright subsists automatically in original works. Therefore, in most countries nothing more is needed to be done in order for a work to attract copyright protection. This remains the case in the US, although there this an additional discretionary ability to register the work with the US Copyright Office.

It’s important to note that registration with the US Copyright Office is distinct from the process of mandatory legal deposit with the US Library of Congress, which is undertaken by Cambridge alongside all other publishers releasing works in the US.

How can there be Cambridge titles on the list, if Cambridge does not register its works with the US Copyright Office?

Individual historical registrations may have been completed, either by Cambridge, individual authors, literary agents, funders or other interested parties, on a case-by-case basis.

If my work is on the list, will I get compensation?

You may have seen various media reports speculating in respect of amounts in compensation to be paid per work infringed.  The reality is that details of how the Settlement Fund will be paid are currently being worked out and will need to be approved by the US Court. Also still to be confirmed is the administrative process for making claims, including whether both Cambridge and the individual author(s) will need to submit requisite forms and information (rather than Cambridge being able to submit for both itself and authors).

Any compensation due to Cambridge authors will be distributed in line with our author contracts, unless any alternative arrangement is directed by the US Court.

If Cambridge has previously reverted rights to you in your work, you will need to make the claim individually – Cambridge would have no grounds to make a claim in respect of your work.

Can I as an author put forward a claim directly in the suit?

If your work is on the list, you may need to submit claim documentation directly. The process will be confirmed in the coming weeks and industry organisations such as Society of Authors and Authors’ Guild may also be able to assist authors in making any direct claims. 

I recently signed the AI addendum to my Author Agreement which was sent by Cambridge. Does this mean I have licensed Anthropic to use my work and now have no claim in this lawsuit?

No, the settlement concerns copyright works which were used by Anthropic without permission or a legitimate licence being in place. The AI addendum is not related to this current Anthropic infringement, nor the resulting settlement. The AI addendum enables Cambridge to license your work for the future training of AI models in legitimate deals that are being negotiated between Cambridge and AI developers. If and when any negotiated AI deal is reached, you will receive royalties of 20% pro rata for usage of your work as set out in the addendum.