Artificial intelligence falls for fake disease, spreads medical misinformation

Dry itchy eyes may indicate an underlying fictitious disease called bixonimania, according to multiple artificial intelligence platforms that have propagated the bogus disease.

Almira Osmanovic Thunström from the University of Gothenburg created the fake condition to test how large language models absorb and promote medical misinformation, according to Breitbart.

On March 15, 2024, Osmanovic Thunström published two blog posts on Medium, describing sore, tired, pink eyes as a developing condition from excessive blue light exposure. She also published two papers on the academic network SciProfiles in late April and May 2024, under the pseudonym Lazljiv Izgubljenovic and an AI-generated photo.

Osmanovic Thunström gave numerous clues the disease was fake, even in its name, according to Nature, which originally broke the story.

“I wanted to be really clear to any physician or any medical staff that this is a made-up condition, because no eye condition would be called mania – that’s a psychiatric term,” she said.

Additionally, throughout the papers she listed fake universities, such as Asteria Horizon University, The Starfleet Academy and the University of Fellowship of the Ring. She even included direct statements discrediting the entire paper’s research: “this entire paper is made up” and “fifty made-up individuals aged between 20 and 50 years were recruited for the exposure group,” Nature reports.

But AI platforms consumed and presented the fraudulent science as fact.

For example, on April 13, 2024, Microsoft Bing’s Copilot called bixonimania a “rare condition,” and Google’s Gemini recommended users with itchy eyes visit an ophthalmologist with concerns of bixonimania, Breitbart reports. That same month, Perplexity AI reported one in 90,000 individuals contracted the disease, and OpenAI’s ChatGPT began diagnosing users, Nature reports.

Eventually, the fake condition made its way into a scientific journal: Cureus, published by Springer Nature. This research cited the fake papers as legitimate sources, marking a failure of the scientific process, according to Alex Ruani, a doctoral researcher in health misinformation at University College London. Cureus retracted the paper March 30, 2026, after Nature informed editors of fraudulent citations.

Nature and Breitbart contacted the AI companies, which either gave no response or claimed AI technology has improved significantly since Osmanovic Thunström’s experiment.

“If the scientific process itself and the systems that support that process are skilled, and they aren’t capturing and filtering out chunks like these, we’re doomed,” Ruani said. “This is a masterclass on how mis- and disinformation operates.”