From Plagiarism to Prompt-Engineering: Nigerian Academic Publishing at the Crossroads of AI Accountability, by Solomon Abiodun Oyeleye

CJID AI and Tech Reporting Fellow
A British journal, aptly titled AI, devoted to publishing peer-reviewed articles on artificial intelligence (AI), has an unusual condition for prospective authors: it does not recognize generative AI tools and Large Language Models (LLMs) as authors, as they don’t meet authorship criteria (https://www.sciencedirect.com/journal/artificial-intelligence/publish/guide-for-authors) .
This reporter sent a prompt to Google Gemini asking for a figurative way to describe the condition spelt out by the AI editors, a journal on AI, telling authors it does not recognize AI as authors. This was Gemini’s response: “It’s like an architectural firm’s journal on modern building tools, declaring that no matter how sophisticated a 3D printer becomes, it can never be credited as a co-designer on the final blueprint.
“The journal is essentially acting as a gatekeeper of consciousness, asserting that a tool — no matter how revolutionary or intelligent it appears — cannot possess the human spark of creativity, ethical responsibility, or original thought required for authorship. It’s a paradoxical statement from an AI journal, essentially saying, “You are the subject of our study, but you are not one of us.”
Retraction Watch, an organization that tracks retractions of academic articles by major publishing firms, has opened a ledger to track such articles withdrawn due to high complaints that they are written by AI. As of the last check in September, there were 92 articles on the list (https://retractionwatch.com/papers-and-peer-reviews-with-evidence-of-chatgpt-writing/,)
In academia, publishing is a core requirement for measuring quality and promotability. Overwhelmed and overworked, lecturers would gladly welcome any assistance towards writing their articles faster and perhaps better.
AI offers that assistance.
However, in the process, quality and authenticity have become suspect.
Anthony Akinwale, Vice Chancellor of Augustine University, Ilara Epe, Lagos State, had, in an earlier interview on AI adoption in universities, expressed concern over what he described as the “other side of the story” while grappling with the adoption. This ‘other side’, according to Akinwale, is the requirement of intellectual honesty, a development the Dominican Friar and Professor of systematic theology and thomistic philosophy, noted “calls for an ethically regulated framework…that AI is not used in a manner that transgresses the boundary of ethics”
Olugbenga Abimbola, a senior lecturer at Adekunle Ajasin University, Akungba Akoko, also in an earlier interview, had recommended two policy thrusts to guide AI use in universities: “For me, AI use in academics is in two aspects: data and literature. For literature, I will propose that we have a policy that will enable us to use AI to edit our works, to proofread and maybe straighten our work, but not to the extent of asking AI to generate content. So, let’s have a policy that will discourage people from using AI to generate literature or academic content. But for the use of data and editing, I will support that.”
While the integration of generative AI tools like ChatGPT into academic writing has transformed scholarly publishing globally and brought various advantages, it has also brought forth many challenges, particularly related to accountability and transparency in the publication process. Hence decision by journals like AI, to not recognise AI as an author, goes to the heart of the matter as the publishing community grapples with the development.
How are Publishers and editors of academic journals coping with this new trend?
It is a source of serious concern for Mike Nwidobie, Professor of Accounting and Editor of Caleb International Journal of Development Studies. “The number of papers generated by AI submitted by authors is increasing rapidly”, he noted during an interview, adding: “The percentage of AI content of a paper generated by AI ranges between 68-100%.”.
Eric Aondover, editor and reviewer for Taylor and Francis Group and SAGE Publications, acknowledged that many authors are still figuring out how to strike a balance between AI writing’s efficiency and the demand for creativity and critical thinking.
“My experience with them has been rather dynamic”, the Communication scholar and senior faculty at Caleb University, Lagos, said.
“Some authors use AI technologies extensively, which frequently leads to well-structured drafts that lack depth, individuality, or appropriate reference. Others employ AI more carefully, using it as a tool to help them organise text or come up with ideas, which facilitates editing. In both situations, I see my job as helping authors polish their writing so that, even if they might benefit from AI-generated content, their finished products will exhibit true scholarship, contextual accuracy, and the distinct viewpoints that only human authors can offer.”
A random sample of several journals published in Nigeria, indicated no attempt to create awareness or provide guidelines on the use of AI in article submission. As at September 18, 2025 when their websites were accessed for this story, the following journals which are still publishing, have not provided guidelines for authors, unlike what the publishers of AI has done: The Nigerian Journal of Pharmacy (https://www.psnnjp.org/index.php/home), published by the Pharmaceutical Society of Nigeria since 1961, Nigerian Research Journal of Engineering and Environmental Sciences (http://rjees.com/,), domiciled in the Faculty of Engineering, University of Benin, Benin City, Edo State, Ibadan Journal of Sociology(https://www.ibadanjournalofsociology.org.ng/,) published by the Department of Sociology, University of Ibadan, African Journal of Educational Management (https://journals.ui.edu.ng/index.php/ajem,) published by the Department of Educational Management, University of Ibadan, since 1983, Lagos Journal of Environmental Studies (https://ljes.unilag.edu.ng/,) published by the Faculty of Environmental Sciences, University of Lagos since 1997, UNILAG Journal of Business(https://ujb.unilag.edu.ng/about,) also published at the University of Lagos by the Department of Business Administration.
In contrast, foreign journals have started inserting guidelines on AI use for their authors. Apart from the journal AI, the Journal of the American Veterinary Medical Association (https://avmajournals.avma.org/view/journals/javma/javma-overview.xml,)has stipulated disclosure demands for authors in relation to AI use. One of such disclosures relates to “the use of any artificial intelligence (AI)–assisted technology such as ChatGPT or another Large Language Model in the writing of the manuscript or production of images.” It is also expected that, if none was used, authors would include the statement “No AI-assisted technologies were used in the generation of this manuscript.”, and if an AI tool was used, the authors must be transparent in disclosing in the disclosures section, which AI tool was used and how the AI tool was used. AI tools cannot be listed as an author of a manuscript. The Canadian Medical Educational Journal (https://journalhosting.ucalgary.ca/index.php/cmej,) has similar provisions.
However, two Nigerian publishers contacted for this story said their journals have made provisions for such awareness. Dr Gever Verlumun Celestine, Editor-in-Chief and Founder, Ianna Journal of Interdisciplinary Studies
(https://iannajournalofinterdisciplinarystudies.com/index.php/1,) said the journal has included a warning against the use of AI to generate manuscript content, while Professor Eserinune McCarty Mojaye, Editor of Journal of Communication and Media Research said his Editorial team members have been trained to decipher where AI has been used by an author.
“We are also aware that some lazy or fraudulent authors may depend solely on AI-generated content. Towards this end, we have sponsored our Editors to workshops and trainings on the use of AI. We have also organized in-house workshops for ourselves on the use of AI, so all our editors are familiar with AI use”, he said in a note to this reporter, hinting that the journal was still in the process of developing its AI policy, “to provide a proper and fair framework and guide” to authors.
How we deal with AI authors
In the absence of clear guidelines, such as the ones provided by the publishers of AI and CMEJ, how are journal editors and reviewers dealing with the suspicious use of AI by authors?
“As a journal editor, I approach the challenges posed by AI-written articles by combining critical analysis with helpful modification”, said Aondover, pointing out that instead of throwing away content that could have been generated with AI, he focuses on “transforming AI outputs into polished, trustworthy, and human-centered work that meets both ethical and professional standards, rather than completely disregarding them”.
However, for Nwidobie, that is not the path to thread: “All submitted articles are subjected to a test of AI content using available software. CIJDS does not accept articles or parts generated by AI. Such articles are usually rejected and returned to the authors. Three or four articles submitted for the forthcoming edition found to be fully or partly generated by AI were rejected.”
For Celestine, his journal has adopted “a proactive and restrictive policy” including the establishment of “clear guidelines for disclosure and are enforcing strict limits on AI use to prevent the intellectual product from being generated by non-human entities. Our focus remains on the prevention and rigorous review of all submissions”.
In a recent article (https://www.timeshighereducation.com/opinion/ai-based-fake-papers-are-new-threat-academic-publishing,)for Times Higher Education, Seongjin Hong, a Professor at Chungnam National University, South Korea, narrated his experience as a journal editor, lamenting that beyond seeking to enhance scholarly publications, there is the possibility that AI generated content are being released to journals to test some developing models without prior information to publishers. The intention may be to find a way to bypass the rigorous peer-review process common to academic publishing.
“As an editor, I am receiving submissions from spurious authors consisting of previously published papers altered by AI. Last month, I noted my strong suspicions that a manuscript of mine had been reviewed by AI. However, it turns out that AI is not only reviewing papers. As the associate editor of a scientific journal, I am also seeing mounting evidence that it is writing papers too – or, rather, that it is plagiarising already published ones.”
We need help…
If lecturers are penalizing students for using AI to write their assignments, what should be done to lecturers who use the same AI to write research articles for publications, to earn promotion and recognition?
Prof. Uwaoma Uche, Deputy Vice Chancellor, Administration at Gregory University, Uturu in Abia state said there is need for universities to step up quality assurance processes in dealing with this emerging trend.
“We know that quite a lot of lecturers are preparing lecture and papers with aid of AI, but in as much as we do that, we also have a Quality Assurance unit in the university that ensures that whenever you use such content there must be an attribution of the source, which is AI, so that people don’t commit plagiarism in the use of AI content because that is a high academic crime.”
Hong said the academic community needs help
“We need help. After all, the consequences of failing to detect AI-written papers are serious. Journals themselves risk damage to their reputations and even removal from citation and other scientific databases, affecting not only the journal but also its authors, editors and affiliated institutions.
“Moreover, academic publications are not just a measure of individual or institutional achievement; they form the foundation for policy decisions and public trust. If fake research slips through and accumulates in the scientific record, the consequences go beyond individual misconduct and threaten the integrity of evidence-based decision-making across society.
“It is now abundantly evident that AI is disrupting the flow of scientific publishing in multiple ways.”
The Committee on Publications Ethics (COPE), a Non -Government organization devoted to promoting ethical standards in academic publishing, recently held a Forum to discuss the dilemma (https://publicationethics.org/topic-discussions/emerging-ai-dilemmas-scholarly-publishing, ), focusing on five critical questions: What levels of disclosure are ethically required from authors regarding their AI interactions? How can AI-use standards and disclosure be enforced? How can publishers create dynamic policies capable of adapting quickly to emerging AI technologies and their implications? How do current AI-detection technologies handle concerns around false positives, biases, transparency, and accuracy? How can publishers create dynamic policies capable of adapting quickly to emerging AI technologies?
Responses are still ongoing, but one of the comments to the questions on how publishers can create dynamic policies to adapt quickly to emerging AI technologies may be a guide for Nigerian academic publishers:
“This is arguably the greatest current failure in academic publishing. Publishers must create AI ethics advisory boards that include scholars, ethicists, and technologists—not just editorial managers. Policies should be iterative, updated biannually if necessary. Editors should be trained to ask better disclosure questions, not just detect or penalize
“The lack of agility among large publishers currently undermines trust in the publishing ecosystem, especially for early-career researchers navigating uncharted expectations.
“What’s needed is a shift from suspicion to collaborative regulation, where AI is not seen as a threat to authorship but as a partner in knowledge construction—one that requires ethical framing, not fear-driven restriction.”
And, like Professor Hong warned: “Failure to adopt such measures will ultimately undermine the foundation of scientific trust.”
This report was produced with support from the Centre for Journalism Innovation and Development (CJID) and Luminate