AI chatbots fronted a fake eye disease and told everyone about it
Bixonimania. At Nature:
It’s the invention of a team led by Almira Osmanovic Thunström, a medical researcher at the University of Gothenburg, Sweden, who dreamt up the skin condition and then uploaded two fake studies about it to a preprint server in early 2024. Osmanovic Thunström carried out this unusual experiment to test whether large language models (LLMs) would swallow the misinformation and then spit it out as reputable health advice. “I wanted to see if I can create a medical condition that did not exist in the database,” she says.
The problem was that the experiment worked too well. Within weeks of her uploading information about the condition, attributed to a fictional author, major artificial-intelligence systems began repeating the invented condition as if it were real.
Chris Stokel-Walker, “Scientists invented a fake disease. AI told people it was real,” April 7, 2026
The red flags were pretty obvious, actually:
The fictional lead author supposedly worked at “Asteria Horizon University” in the nonexistent “Nova City, California.”
The acknowledgements section thanked “Professor Maria Bohm at The Starfleet Academy” on the USS Enterprise, with funding credited to “the Professor Sideshow Bob Foundation for its work in advanced trickery.”
The papers even stated outright that “this entire paper is made up.”
Thunström chose the name bixonimania deliberately. The suffix “-mania” is used exclusively in psychiatry, so no legitimate eye condition would ever carry that label.
Angelina Walker, “‘Bixonimania’ Is a Fake Disease—But ChatGPT Diagnosed It to Thousands, Other AI Did Too,” Nurse, April 10, 2026
It’s not a problem unique to invented diseases, say at Susan A. Nolan, Ph.D., and Michael Kimball Psychology Today:
How bad is the problem of AI inaccuracy in medicine? Researchers recently asked five AI chatbots 10 questions from a range of medical areas (Tiller et al., 2026). The chatbots produced “problematic” responses about half the time, and about 20% of all responses were “highly problematic.” “Do You Have Bixonimania?,” May 15, 2026.
We should keep this in mind when we hear commentators extol the wonders of AI in health care.
