Scientists are horrified by a strange AI mouse with huge reproductive organs in a peer-reviewed article

Horror and disdain swept through scientists' social networks on Thursday, as several terribly bad numbers generated by artificial intelligence circulated from… Peer-reviewed article Recently published in a prestigious magazine. These numbers – which the authors acknowledge in the text of the article are Midgourney's creation – are all inexplicable. They contain incomprehensible text, and most strikingly, one of them includes a picture of a rat with strangely large and strange genitalia, as well as the text title “dck.”

On Thursday, the review article's publisher, Frontiers, Posted an “Expression of Concern” Noting that she is aware of the concerns related to the published piece. “An investigation is currently underway and this notice will be updated accordingly after the investigation is completed,” the publisher wrote.

The article in question is titled “Cellular functions of spermatogenic stem cells in relation to the JAK/STAT signaling pathway,” and was written by three researchers in China, including corresponding author Dingjun Hao from Xi'an Honghui Hospital. It was published online Tuesday in the journal Frontiers in Cell and Developmental Biology.

Frontiers did not immediately respond to Ars' request for comment, but we will update this post with any response.

The first figure in the paper, the one containing a mouse, attracted immediate attention as scientists began sharing it widely and commenting on it on social media platforms, including Bluesky and the platform formerly known as Twitter. From a distance, it is clear that the anatomical picture is completely wrong. But looking closer reveals more flaws, including the designations “dissilced,” “stem cell,” “iollotte sserotgomar,” and “dck.” Many researchers have expressed surprise and dismay that such a blatantly bad image generated by artificial intelligence could pass through the peer. – The review system and any internal processing present in the journal.

But the rat pack isn't the only problem. Figure 2 is less graphic but equally distorted. While it's supposed to be a diagram of a complex signaling pathway, it's instead a jumbled mess. An expert in scientific integrity I wondered if she was giving an overly complicated explanation of “how to make a donut with colorful sprinkles.” Like the first image, the diagram is full of nonsense text and confusing images. Figure 3 is no better, presenting a collection of small circular images containing gibberish-heavy annotations. The image is supposed to provide visual representations of how the signaling pathway from Figure 2 regulates the biological properties of spermatogenic stem cells.

Some scholars online wondered whether the text had been generated by artificial intelligence as well. One user noted that AI detection software determined it was likely generated by AI; However, as Ars previously reported, such programs are unreliable.

The images, although egregious examples, highlight a growing problem in scholarly publishing. The success of scientists depends largely on their publication record, with a high volume of publications, frequent publication, and articles appearing in top journals all earning scientists greater status. The system incentivizes less scrupulous researchers to publish low-quality articles, which are likely to be generated with the help of artificial intelligence in the age of AI-powered chatbots. Researchers are concerned that the increasing use of artificial intelligence will make published research less trustworthy. As such, research journals have recently developed new guidelines for authoring AI-generated texts to try to address the problem. But for now, as the Frontiers article shows, there are clearly some gaps.