AI & Authorship
Writing about web page https://publicationethics.org/cope-position-statements/ai-author
Like many of you I've been following the discussions around authorship and AI, especially as it relates to ChatGPT and scholarly communications in recent weeks (for example: Haggart, 2023; Lucey & Dowling, 2023). You probably saw the splash in the news recently too when the major research journal Science took the position ‘banning the use of text from ChatGPT and clarifying that the program could not be listed as an author.’ (Sample, 2023). Naturally, as a journal editor thoughts on originality are rarely far from my mind, and while there have been tools around for some time which can be deployed by authors in the creation of text – I can recall playing with them as far back as two decades ago – ChatGPT does rather seem to have shifted the practice from a niche to a mainstream activity.
Given Exchanges regularly looks to COPE (the Committee in Publishing Ethics) for best practice guidance in maters of publishing ethics, I’ve been keeping our powder dry as far as any related policy for the journal is concerned. Certainly, during the last few weeks we’ve received our first – and I doubt last – article submission relating to the issue. Note about not by, being the important elements in this respect. Nevertheless, I suspect in the fullness of time we may will almost certainly have contributions from authors who will be making use of AI tools in the creations of their papers.
Hence, this morning I noted with particular interest how COPE have now produced a position statement on the issue of authorship and the use of AI in the creation of research publications. I confess I’ve been waiting on this with anticipation, and now it’s here am glad to report it is fairly elegant in its simplicity. The key elements of COPE’s position being:
- AI Tools cannot be listed as paper authors given they cannot take any legal responsibility.
- Authors utilising AI tools in a manuscript’s creation must disclose where/how they were used.
- Authors retain responsibility and ethical liability for all of their paper’s contents
To my thinking this seems a rational, fair and workable approach. It doesn’t entirely exclude contributions from authors who may well wish to make a use of AI tools in the creation of their research outputs – which is good, because I wouldn’t want to preclude these from our considerations. However, it does clarify and demarcate the expectations of professional ethics, original contributions and the boundaries of authorship within any journal contributions. While I suspect questions around the use and misuse of AI tools within scholarship will not be evaporating any time soon, to my mind this position statement at least provides editors like myself with a framework upon which to consider and build our own related submission policies.
As such, with the hopeful agreement of Exchanges’ Editorial Board, we will be adapting and adopting a policy on AI tools and authorship, very much based on the COPE guidance. Given we as a journal typically look towards COPE for best ethical practice, this is line with the development of our extant policy frameworks. Authors seeking to explore these topics more, and how it may impact on the production of their submissions, naturally, are encouraged to contact myself for further discussions.
COPE, 2023. Authorship and AI Tools: COPE Position Statement. Committee on Publication Ethics. https://publicationethics.org/cope-position-statements/ai-author
Haggart, B., 2023. ChatGPT Strikes at the Heart of the Scientific World View. Centre for International Governance Innovation. 23 January 2023. https://www.cigionline.org/articles/chatgpt-strikes-at-the-heart-of-the-scientific-world-view/
Lucey, B., & Dowling, M., 2023. ChatGPT: our study shows AI can produce academic papers good enough for journals – just as some ban it. The Conversation, 26 January. https://theconversation.com/chatgpt-our-study-shows-ai-can-produce-academic-papers-good-enough-for-journals-just-as-some-ban-it-197762
Sample, I., 2023. Science journals ban listing of ChatGPT as co-author on papers. The Guardian, 26 Jan 2023. https://www.theguardian.com/science/2023/jan/26/science-journals-ban-listing-of-chatgpt-as-co-author-on-papers