Skip to main content

Artificial Intelligence policy

With the rapid development of artificial intelligence (AI) tools and technologies, and their increasing availability to researchers, these guidelines have been produced to provide support and clarity for authors and editors on the acceptable use of AI in the preparation of University of London Press (hereafter UoL Press) publications, and requirements for disclosure.  

This policy primarily applies to the use of generative AI tools. Generative AI refers to systems that can generate content including text, images, video, audio and translations in response to users’ prompts. These tools are trained on large datasets and use machine learning to identify patterns in order to generate content. Large Language Models (LLMs) are types of generative AI tools and examples include ChatGPT, Gemini, Copilot and Claude. 

UoL Press permits limited, responsible use of generative AI tools in the preparation of publications by authors. By responsible, we mean that the AI use is transparent, respects data protection, confidentiality, privacy and intellectual property rights, and that the accuracy of the work and its originality to the author are maintained. 

While we recognise that generative AI can support and enhance elements of the research and writing process, these tools still currently have significant shortcomings, for example, in reliability, data security, transparency, the introduction of biases and their environmental and human impacts (see the University of London’s Considerations for Responsible AI guidelines). It is therefore essential that any use of these tools by authors is responsible and judicious, and our purpose to publish work that reflects the original scholarship, creativity, analysis, conclusions and voice of the author remains unchanged.   

Policy overview: requirements of authors 

Book proposals (with their accompanying sample chapters) and final submitted book manuscripts are in the scope of this policy. Authors are required to: 

  1. Be accountable for all intellectual content in their book proposal or manuscript. Authors remain fully responsible for all material, including content produced with the assistance of generative AI in one of the acceptable ways listed in this policy. Authors must verify that the work is original to them. 
  1. Review all use of generative AI against this policy and adhere to it. Unacceptable uses of generative AI in the writing and manuscript preparation process outlined here must be avoided. If using generative AI tools as part of the research process itself, you should also adhere to your institutional and/or disciplinary guidance on the ethical use of generative AI in research. 
  1. Ensure that human oversight of any AI-generated content is maintained. All content generated by an AI tool must be reviewed and checked for accuracy by authors as a vital part of having accountability and responsibility for the content of their submissions. Generative AI can introduce factual inaccuracies, fabrications (or ‘hallucinations’), biases and unattributed use of third-party content, so all use must be carefully checked. 
  1. Disclose all use of generative AI in the book proposal or manuscript preparation process on submission (except where otherwise indicated). Where generative AI tools have been used, transparent declaration of how they’ve been used and whereabouts in your proposal or manuscript is essential to ensure your book meets our standards of ethical publishing. Guidance on how to disclose is provided in this policy. 

Acceptable and unacceptable uses of generative AI 

Generative AI may only be used as an aid for elements of the writing process, and never as a replacement for the author’s work. Any use of generative AI that amounts to fabrication (making up content or results), falsification (misrepresenting data or results) or plagiarism (using another’s work without credit) in the research or the writing is not permitted, as these are serious breaches of research integrity and ethical standards in publishing. 

Acceptable uses 

Authors may use generative AI tools in the preparation of their proposal or manuscript to: 

  • Check and refine their own original writing, for example, by using spellcheckers and grammar tools. Disclosure not required. 
  • Format references. Disclosure not required. 
  • Translate short quotations to or from English or another language. Disclosure required as well as review by a speaker of the language(s). 
  • Present or format their research data. Disclosure required. 
  • Create alt text for images that are the author’s own or in the public domain (images under third-party copyright must not be uploaded to AI tools). Disclosure required. 

Always check the terms of use of the AI tool to ensure it offers sufficient protection around data security and confidentiality. Use a private or institutional account where possible. 

Unacceptable uses 

Authors must not use generative AI tools in the preparation of their proposal or manuscript to: 

  • Generate text for inclusion in the work as a replacement for the author’s own writing (as distinct from checking and refining their own original writing). 
  • Generate research arguments or interpretations that are passed off as the author’s own analyses. 
  • Translate full chapters of their book to or from English or another language except with express written agreement from UoL Press’s editorial team. 
  • Generate or manipulate images, other than in exceptional cases with express written agreement from UoL Press’s editorial team (see below for more information). 
  • Produce references or reference lists. All sources cited must have been consulted and verified by the author, and be relevant to the research. 
  • Generate data (fabrication) or alter existing data or results such that the research is not accurately represented (falsification). 

These lists are not comprehensive and your planned use of an AI tool may not be listed here. If you are unsure whether your intended use of generative AI is acceptable then please check with your editorial contact at UoL Press, ideally prior to use, and ensure you disclose it on submission.  

Authors are also expected to follow their institution’s or discipline’s research ethics guidance on acceptable and unacceptable uses of generative AI as part of their research methods and describe any use in their methodology. 

Disclosure 

Authors are required to disclose all generative AI use in their proposal form and manuscript submission form (except where specifically exempted in this policy), using the fields provided to do so. Any use of these tools as part of the research process itself should also be declared in a methodology section in your manuscript. 

You should disclose the name of the generative AI tool, the model/version used (where available), the purpose for which it was used and the sections of the manuscript or proposal to which it contributed. 

We will use the information you disclose to offer guidance on whether the use is acceptable under our policy and whether further disclosure in the text of your manuscript is required. It will also help to improve our own understanding at UoL Press, as we learn more about how researchers are making use of generative AI. Our editorial team may contact you with questions about AI use in your submission, whether disclosed or undisclosed. 

Failure to disclose material use of generative AI may constitute a breach of publishing ethics and may result in rejection, withdrawal of publication, publication correction or other action. 

We would strongly recommend that you document any generative AI use while you are working on your manuscript to make the process of disclosure easier at the point of submission. 

Further guidance for authors 

Authorship 

Authors must be accountable and responsible for all intellectual content in their manuscript. In line with COPE’s position statement on Authorship and AI Tools (2023), AI cannot be listed as an author or co-author on our publications as it is not able to take responsibility for the work, enter into licensing agreements nor declare any conflicts of interest. 

Copyright and open access  

All open access UoL Press books are published under Creative Commons (CC) licences, which are licences applied by the copyright holders (i.e. the author, in the case of the books we publish) that enable others to reuse their work under a standard set of terms. The terms of the CC licence applied to your book should be respected by AI providers – for example, UoL Press’s preferred CC BY-NC-ND licence should prevent the use of the book by AI for commercial purposes and in a derivative form. However, please note: 

  • Attribution (the ‘BY’ in a CC licence) is not always possible when a generative AI tool returns its results, as they use massive underpinning datasets combined from a range of sources to generate content. 
  • Text and data mining exceptions in UK legislation may allow for the training of AI systems on copyrighted works such as books. 

There is substantial ongoing discussion and development of copyright law across jurisdictions to legislate on how AI providers may use copyrighted works (whether open access or not), so this guidance will be kept under review. 

Useful links: 

More information on the relationship between the Creative Commons and AI can be found on the Creative Commons website.  

Images 

We do not permit the inclusion of images that have been created or manipulated using generative AI. This is for a number of reasons, for example, the fabrication or distortion of images used as evidence in your book (e.g. historical images to support claims) can amount to a breach in research integrity. The AI tool’s potential use of third-party imagery to generate its own images is also not transparent so using that image could infringe third-party rights. 

An exception to our image policy can be made if you wish to use generative AI to present a simple graph or diagram based on your own research data or where generated content forms part of AI-related research itself, though you must disclose this use and it will be agreed on a case by case basis by UoL Press’s editorial team. 

Edited collections 

Volume editors should make the contributing authors to their edited collection aware of the requirements of them regarding generative AI use under this policy. 

Editorial evaluation of contributors’ chapters should always be carried out by the editors without the use of generative AI. Editors are responsible for their own editorial decision-making. 

If editors have any concerns about inappropriate use of generative AI by their contributors they must raise these with UoL Press’s editorial team for support. 

Uploading your content to generative AI tools 

Before uploading any of your work to a generative AI tool, ensure to check its terms for details of any rights you are granting to the AI provider that might impact your ability to enter into a publishing agreement with us or to apply a Creative Commons licence to your work. We would also recommend you check that you are not giving them any additional rights with which you are personally not comfortable. 

There are known information security and privacy issues with using public generative AI tools so it is important that you remove any confidential (including legally privileged), personal or sensitive data from your material before uploading it. 

AI guidance for peer reviewers 

Peer reviewers must not upload authors’ proposals or manuscripts (or sections from them) that they’ve been asked to review to a generative AI tool, as this risks infringing the authors’ copyright in their work and may also involve sharing confidential or personal information.  

We also do not permit peer reviewers to use generative AI to produce a written peer review of the proposal or manuscript. Peer reviewers are responsible and accountable for their own evaluation of, and report on, an author’s submission. If required, peer reviewers may use generative AI tools to assist with checking and refining their language, spelling and grammar in their report. 

How is UoL Press using AI?  

We do not use generative AI tools in the editorial evaluation of new proposals and contracted book manuscripts submitted to us, and work to ensure that our production processes also limit the use of AI. As we work with a range of different suppliers, vendors and platforms across our publishing workflow it is possible that they may make changes in their working practice to introduce generative AI use. We will take steps to ensure that any use is known to us and will update these guidelines to reflect any changes transparently. 

We have not signed any licensing agreements with AI providers to allow them to use UoL Press books to train AI models. 

We have started to experiment with some generative AI use in our marketing activities and to explore how AI tools might support increased access to our books. This is still at an early stage and we will share updates as our use of AI develops.   

Useful links 

This policy will be reviewed every 6 months. Last updated 29 July 2026.