The AI-Assisted Academic: A Tale of Questionable Productivity
A prolific academic's publications and reviews are under scrutiny, sparking a debate on the ethics of AI usage.
A neuroscientist in Germany stumbled upon a peculiar manuscript, raising questions about the integrity of its content. The paper, authored by Eren Öğüt, an associate professor at Istanbul Medeniyet University, aimed to model a brain structure called the indusium griseum. But the reviewer was perplexed by the lack of clarity and the seemingly irrelevant MATLAB functions. The figures were unintelligible, and the discussion resembled a literature review, failing to connect with the results.
The reviewer's suspicion grew when they questioned the MRI data's resolution, which seemed inadequate for visualizing the intricate brain structure. Consulting a colleague, an expert in brain imaging, confirmed their doubts. The resolution was indeed too low for the task.
But the story takes a twist when the same paper, with minor changes, resurfaces for review at Scientific Reports. The second reviewer, intrigued, found the manuscript filled with scientific jargon but lacking substance. The figure, meant to illustrate the indusium griseum, displayed a simple sinus wave, raising doubts about its authenticity.
Here's where it gets controversial: Öğüt's extraordinary productivity has raised eyebrows. In 2025 alone, he published 25 papers, almost exclusively in Springer Nature journals, with 12 being single-authored. Astonishingly, he also reviewed nearly 650 papers that year, with over 1,400 total reviews to his name. How is this level of output possible?
The answer, according to the reviewers, might lie in the reckless use of generative AI. Öğüt's reviews, averaging 364 words, align suspiciously with the average review length from 11 million reviews. The reviewers believe this could be a strategy to enhance his resume and gain favor with journal editors.
Öğüt, however, defends his work, attributing his productivity to team effort and claiming AI is only used for editing and improving sentence clarity. He emphasizes his commitment to meeting deadlines and working with a dedicated team.
But the plot thickens. After being contacted, Öğüt's online profiles on Google Scholar, ORCID, and Frontiers' Loop mysteriously disappeared.
The reviewers' concerns reached John Van Horn, editor of Neuroinformatics, where three of Öğüt's single-authored papers were published in 2025. They noted a pattern of similar templates, redundant figures, and discussions that resembled literature reviews. The author never provided real MRI images, shared data or code, and even stated that no datasets were generated or analyzed.
Springer Nature's Research Integrity Group is now investigating Öğüt's submissions to Neuroinformatics. While the investigation is ongoing, the head of Research Integrity assures the matter is taken seriously.
One of Öğüt's papers, published in Neuroinformatics, claims to use a method developed by mathematician Anuj Srivastava and colleagues. Strangely, it presents a specific quantitative result (38%) as a finding from Srivastava's work, which is not mentioned in the original article. Öğüt clarifies this as an approximation, but the paper's quality is questioned by Srivastava and co-author Jennifer S. Stevens, who find it vague and lacking methodological details.
This case raises important questions: How did these papers pass peer review? Is the use of AI in academic writing ethical? And what does it mean for the integrity of scientific research? The debate is open, and we invite our readers to share their thoughts in the comments.
