Academic Integrity in Question: AI-Generated Papers or Human Expertise?
A neuroscientist reviewer's confusion sparks an investigation into a prolific academic's work.
Imagine a manuscript that leaves you scratching your head, questioning if it's a masterpiece or a machine's musings. This is the dilemma faced by a German neuroscientist reviewing a paper on modeling the indusium griseum, a brain structure. The article, authored by Eren Öğüt, an associate professor at Istanbul Medeniyet University, seemed peculiar. The figures were unclear, the provided MATLAB functions irrelevant, and the discussion lacked engagement with the results.
But here's where it gets controversial: the reviewer, suspecting low-resolution MRI data, consulted a colleague, who confirmed her doubts. Despite this, Öğüt's paper found its way to another journal, Scientific Reports, where a second reviewer encountered similar issues. The text seemed like AI-generated technobabble, and the figures were baffling.
And this is the part most people miss: Öğüt's productivity is astonishing. In 2025, he published 25 papers, almost exclusively in Springer Nature journals, with 12 being single-authored. He also reviewed nearly 650 papers that year, according to Clarivate's Web of Science. His average review length of 364 words is suspiciously close to the average of 11 million reviews.
Öğüt defended his work, attributing his productivity to teamwork and ongoing research. He admitted to using AI tools for editing but denied any unethical practices. However, his online profiles mysteriously disappeared after being contacted.
The reviewers' concerns reached John Van Horn, editor of Neuroinformatics, who confirmed their suspicions. Öğüt's single-authored papers followed a similar pattern, with redundant figures, irrelevant MATLAB functions, and discussions resembling literature reviews. The author failed to provide real MRI images, data, or code, raising further doubts.
The investigation is now underway at Springer Nature, with the Research Integrity Group examining Öğüt's submissions. Tim Kersjes, head of Research Integrity, assures the matter's seriousness. But the question remains: how did these papers pass peer review?
One paper, published in Neuroinformatics, claims to use a method by mathematician Anuj Srivastava but provides vague details and misrepresents Srivastava's findings. Srivastava and co-author Jennifer S. Stevens find Öğüt's work subpar and confusing.
This case highlights the challenges of maintaining academic integrity in the age of AI. Is Öğüt's productivity a testament to human expertise or AI assistance? Should journals reconsider their peer review processes? Share your thoughts below, and let's spark a discussion on the future of academic publishing.
