Nothing was alarming until Dan Quintana, a professor of psychology at the University of Oslo, looked at a reference in a paper he was reviewing and saw his own name. The reference to his work looked correct—it had a plausible title and included authors he had worked with in the past—but the article to which the work referred did not exist. Much more worrying is the ability of generative AI to conjure up convincing pictures of thinly sliced tissue, microscopic fields, or electrophoresis gels that are commonly used as evidence in biomedical research.

Pangram Labs recently analyzed thousands of peer reviews that were submitted to ICLR, the leading conference for deep learning, and found that more than half of them were written with help from an LLM, and about a fifth of them were wholly AI-generated.

Unfair use of AI can lead to retractions of articles and damage the reputation of a scientist. AI is a useful tool, but it is not a universal solution for all stages of research or article writing. Saving time is not worth losing trust. Be careful, learn to use AI responsibly – watch the video "Why authors should not rely on artificial intelligence when preparing a reference list" and "AI and science: why аrticles are being massively removed from scientific databases".

Source https://www.theatlantic.com/science/2026/01/ai-slop-science-publishing/685704/