The European Journal of Development Research (EJDR): a note on papers submitted to the journal, and the use of generative AIs such as Large Language Models (LLMs)

The researcher and academic community is increasingly vigilant about the growing use of Generative Artificial Intelligence (AI) tools and Large Language Models (LLMs; such as ChatGPT) in academic knowledge production. In parallel, there is an understanding that such tools are here to stay and will become more pervasive.

EJDR believes that a healthy debate around the use of such AIs in academic research and publication, resulting in a firm set of guidelines, is necessary. Since it is still early days, and since a mature debate around the use of AI and LLMs still needs to be finalised, EJDR has drafted its own guidelines for authors around the use of generative AIs in submissions to this journal. Please note that these guidelines will be periodically revised to take into account the latest developments and industry position.

EJDR’s guideline on Generative AI are aligned with those of the journal’s publisher, Springer Nature under the Palgrave Macmillan imprint:

  • AI or AI-assisted tools cannot qualify as authors, only humans can.
  • Authors are fully responsible for the entire content of their work.
  • Authors must correctly label and disclose which parts of their work have been created by or in assistance with Generative AI.
  • Reviewers and Editors are obliged to confidentiality and should not upload manuscripts to software or AI-assisted tools (and where confidentiality cannot be assured).

EJDR’s instructions for authors around the use of generative AIs are as follows:

  • EJDR does not currently endorse the use of generative AIs in papers submitted to the journal.

There is an understanding that AI can increase productivity by automating micro-tasks and assisting with some aspects of research. Other areas in which the use of AI appears to be growing are ideation and feedback, writing, background research, data analysis, coding, and mathematical derivations1.

However, EJDR believes that submissions to the journal, based on the work and ingenuity of researchers, academics, and practitioners across the international development community, will result diminished if the nuts and bolts of such work – from drafting text to the analysis of data – is outsourced to Generative AI. The formative processes whereby early career researchers learn the keys to academic publication cannot be substituted by farming these processes out to a third-party intelligence, just as learning cannot be obviated or superseded by the use of LLMs.

In alignment with journals such as Nature, EJDR believes that “research must have transparency in methods, and integrity and truth from authors”2. The use of AI-generated content in some academic articles, versus others that have not made use of these tools, risks creating an unlevel playing field and undermining the work of researchers themselves.

EJDR thereby encourages all its authors to make use of their ingenuity and drive to continue producing excellent articles which are, through and through, the fruit of their labour and originality.

  • AI tools or LLMs or similar cannot be listed as authors or co-authors on any papers submitted to EJDR.

Authors are responsible and accountable for the content of their papers, which LLMs are not: they do not have a juridic persona, and their use in any form to provide content for an EJDR submission is strictly under the supervision and responsibility of the paper’s human authors.

Therefore, all authors listed on an EJDR submission must be human; EJDR guidelines on authorship and its interface with AI are aligned with COPE (Committee on Publishing Ethics)3.

  • Any eventual (and sparing) use of Generative AIs must be declared in the acknowledgments and methodology sections of papers submitted to EJDR.

If a paper submitted to EJDR has made use of Generative AI in any form, this use must be sparing and well-justified; any overuse or abuse will automatically disqualify submissions for consideration in the journal. For example: was an LLM/ AI tool used in the collection and analysis of data, and if so, why? Did authors encounter resource issues in the course of their study, which rendered recourse to alternative (namely, AI-based) means necessary and indispensable?

Any use of Generative AI/ LLMs must be acknowledged in two ways: firstly, in the acknowledgments section of the paper; and secondly, in detail, in the Methods section of the paper. Please note that, in addition to this being an EJDR requirement, this is also a COPE and Palgrave Macmillan (EJDR’s publisher) recommendation.

Submission checks are carried out on all manuscripts submitted to EJDR. Particular care is given to plagiarism and reference checks, the latter particularly relevant in the case of LLMs use.

Any queries related to the points above can be addressed to EJDR’s Managing Editor, Talia Vela-Eiden, at vela-eiden@eadi.org

For general queries about journal authorship, please visit Journal Authors | Palgrave Macmillan | Springer

EJDR’s instructions for reviewers around the use of generative AIs are as follows:

In line with the instructions for authors above and those of the publisher, EJDR does not encourage the use of Generative AIs or LLM by reviewers invited to comment on papers submitted to the journal.