News and Events
Uncharted territory: AI and the Cambridge approach for academic book publishing
There has been much debate recently relating to the licensing of academic book content to providers of generative artificial intelligence (Gen AI) tools. This is uncharted territory for publishers and authors, and difficult to navigate as a result. We want to share some of the thoughts behind our approach to AI for academic book publishing at Cambridge University Press.
We know that Gen AI technologies offer both significant opportunities and risks. Some of these are evident already while others may be forthcoming and hard to predict. Gen AI technologies are still emerging, and the nuances are quite distinct from other derivative licensing rights. Our authors and editors are our partners on this journey, which is why working collaboratively and transparently to understand their views is the right approach for Cambridge. For our book authors this means we are actively seeking consent.
"While it could be argued that AI is just another form of sublicensing, and therefore covered by existing rights clauses in our contracts, we felt that it was important to recognise the fundamental differences around potential new content generation and ownership inherent to these deals, and also apprise our authors of some of the principles that we are developing to ensure their books are used and attributed fairly."
Ben Denne,
Director of Publishing, Cambridge University Press
Gen AI tools are likely to become an important component of how information is shared and discovered, and access to high-quality academic content may contribute to increasing the quality of these tools. Engaging constructively with Gen AI providers means we can ensure authors rights are considered as these tools are developed. Encouraging and collaborating in lawful routes to content licensing should help to ensure fair remuneration for the use of authors’ work and allow us to set legal parameters on that use. In many cases, gen AI tools are being built that increase the discoverability, and therefore the impact, of the work we publish, or being used as a component of existing licensing deals that are important for providing access to the content. These good reasons to participate do come with uncharted complexities in how best to govern and limit the different uses of the content we publish, and the technology continues to evolve at pace. Critically, before we proceed with any licensing deals, we want to understand how providers will protect our authors’ moral rights in their work, such as their right to accurate attribution.
Given this balance of opportunity and complexity, we are getting in touch with our book authors and editors of appropriate published titles and establishing their specific consent prior to including their works in any gen AI licensing deals. We want to share our thinking with our authoring partners and give them the opportunity to consider this new area of licensing before deciding to opt in. As we continue to work through this process, we have been pleased to see that many authors, after considering the current context and our approach, have decided to opt in, allowing us to move forward with licensing conversations.
“In relation to the licensing arrangements themselves, clearly there is a significant opportunity not just for a new revenue stream for rightsholders, but also for discovery and collaboration in the development of an exciting new technology and its applications to research and learning. We need to balance the quickly-changing nature of this area with the need to start pinning down how Gen AI licensing agreements can operate effectively. To that end, we worked to establish a set of principles that are guiding our negotiations with gen AI providers, and that we are planning to carry through into individual agreements as appropriate to the use that will be made of the work. For example, we have been emphasising the importance of accurate citation in our conversations with providers. These principles are a guide and we will continue evaluating each licensing arrangement for its specific requirements and scope, as our collective knowledge of this area grows and changes.”
Chris Bennett,
Global Commercial Director, Cambridge University Press
These conversations are also emerging in the world of journals, and come with different considerations and complexities. The importance of establishing the right principles is just as key there and needs to be well thought out and carefully considered and discussed with our society partners.
We remain open to thoughts on our approach from the academic community and will continue to work with our authors to collaborate on the best way forward as Gen AI technologies and their implications for publishing continue to emerge.
