Recent reporting from the respected Dutch newspaper de Volkskrant highlights a growing problem in scholarly publishing: the rise of AI-generated fake citations appearing in academic papers.
According to the article, authored by Stan van Pelt, fabricated references — studies, articles, or journals that do not actually exist — have increased significantly alongside the widespread adoption of generative AI tools. Researchers have found that some AI systems produce highly convincing citations complete with plausible titles, author names, and journal formatting, despite the sources being entirely fictional.
The concern is especially pronounced in biomedical research, where inaccurate citations can influence future studies and potentially affect clinical understanding. Experts interviewed by de Volkskrant warned that the pressure to publish quickly, combined with overwhelmed editorial and peer-review systems, creates conditions in which these false references can slip into published literature unnoticed.
In the article, Simon Linacre, Chief Commercial Officer at Cabells, notes that this issue “shows how great the temptations are for researchers,” and that the danger is not simply that fake references exist, but that they undermine confidence in the integrity of the research itself.
As AI becomes increasingly integrated into academic workflows, the reporting from de Volkskrant serves as an important reminder that efficiency and innovation must be balanced with rigorous verification and oversight. As Simon notes in the piece, “that process is clearly not foolproof.” Maintaining trust in scholarly communication will require researchers, publishers, and institutions alike to remain vigilant as these technologies continue to evolve.
Original article, de Volkskrant: “Aanzienlijke toename van wetenschappelijke verwijzingen naar nepbronnen door AI“
Image credits: de Volkskrant and Raymond Rutting
