Swedish Researcher Exposes AI Chatbots' Vulnerability by Creating Fake Eye Disease

A Swedish AI researcher successfully fooled artificial intelligence chatbots into promoting a non-existent eye condition called "bixonomania," highlighting serious concerns about AI-generated medical misinformation. Almira Osmanovic Thunström from the University of Gothenburg created the fictional disease, supposedly causing red eyelids from excessive screen time, to test how AI systems handle false medical information. The chatbots not only accepted the fake condition as real but began recommending it to users seeking medical advice. "People have stopped being critical about the sources from which they get information," Thunström said in a Science Quickly podcast by Scientific American magazine. The experiment demonstrates the potential dangers of relying on AI chatbots for medical guidance, as these systems can perpetuate and spread false health information without proper verification mechanisms.