In a striking experiment, Almira Osmanovic Thunström, a researcher at the University of Gothenburg, created a fictitious eye condition named bixonimania to test how artificial intelligence (AI) can spread medical misinformation. The condition, purportedly caused by excessive screen time, was entirely fabricated and presented in two preprints uploaded in 2024 under a fictional researcher's name. Despite clear indications of its fictitious nature, including humorous funding sources and acknowledgments, major AI models such as Google's Gemini and Microsoft's Copilot accepted bixonimania as a real condition, with Google suggesting it was linked to blue light exposure and Perplexity AI estimating its prevalence at one in 90,000 individuals. Alarmingly, these fake papers were cited in a legitimate peer-reviewed journal, Cureus, until their retraction in March 2026, following inquiries from Nature. Thunström's experiment serves as a crucial reminder of the potential dangers of relying on AI for medical information, as these systems may confidently provide inaccurate diagnoses. Patients are encouraged to consult qualified healthcare professionals and verify any medical information from trusted sources, as AI cannot replace the expertise of a real doctor.