Generative AI tools now produce text that is increasingly difficult to distinguish from human writing. They can draft literature reviews, polish prose, summarise data and suggest arguments. For researchers under pressure to publish, the temptation is obvious; for editors and reviewers, the problem of detection is growing.
The questions are harder than they first appear. Is text drafted by AI and revised by an author plagiarism? Who is accountable for an error introduced by a tool? Should AI assistance be disclosed, and in what detail? Can a reviewer use AI to assess a manuscript without breaching confidentiality? Journals and publishers have begun to answer, but inconsistently, and many researchers are unsure what is permitted.
Whether AI reshapes how research is disseminated or reinforces traditional methods remains uncertain. It could lower barriers for researchers writing in a second language, speed up routine work and widen access to scholarship. It could also flood journals with low value submissions, weaken peer review and erode trust in the published record.
What is certain is the need for clear guidelines that ensure AI strengthens, rather than undermines, the integrity and credibility of scholarly publication. Those guidelines should be practical: disclosure of how AI was used, human accountability for every claim, protection of confidential review, and attention to equity, so that rules written for well resourced researchers do not disadvantage others.
Above all, the academic community should lead this conversation rather than react to it. Researchers, editors and universities understand what scholarship requires; if they do not set the norms, technology companies and market pressure will set them instead.
Context
He writes from inside the system he describes: as an editor in international journals, Editor-in-Chief of the Book of Future Environments Research and lead editor of conference proceedings with Springer Nature, and as a member of the reference group for the Prime Minister’s Chief Science Advisor’s work on artificial intelligence in healthcare.