AI Chatbots Give Dangerous Sleep Apnea Advice in UK Study

AI chatbots are steering sleep apnea patients away from medical care

Artificial intelligence tools designed to answer health questions are giving potentially harmful guidance to people experiencing sleep apnea symptoms, according to new research from the United Kingdom.

In approximately one-third of interactions studied, AI chatbots incorrectly told patients their sleep apnea symptoms were not serious enough to warrant concern. Medical experts warn this pattern could actively discourage people from consulting specialists about a condition that requires professional diagnosis and treatment.

The research team identified a particularly troubling dynamic: when patients already feel hesitant about seeing a doctor, AI systems tend to validate that reluctance rather than encourage appropriate medical consultation. The technology appears to reinforce existing biases rather than provide objective medical guidance.

Why it matters

Sleep apnea is a serious condition where breathing repeatedly stops and starts during sleep, potentially leading to cardiovascular problems, daytime fatigue, and other health complications if left untreated. Misdiagnosis or delayed treatment can have significant health consequences. As more people turn to AI chatbots for preliminary health information, the accuracy and safety of these systems becomes a critical concern for medical professionals and technology developers alike. The findings suggest current AI health tools may lack the nuance and caution necessary for triage decisions.

The broader context of AI in healthcare

This sleep apnea research emerges as AI systems increasingly intersect with personal health decisions. The chatbot findings highlight fundamental questions about how these tools handle medical uncertainty and risk assessment.

The study underscores a gap between AI capabilities and the complexity of medical evaluation. While chatbots can process symptoms and provide information, they may not adequately account for the serious consequences of missed diagnoses or the importance of professional examination for conditions like sleep apnea, which often require sleep studies and specialized equipment for proper diagnosis.

Additional health technology findings

The same report from WDBJ7 included other health-related research developments. A Stanford Medicine study found that warning labels about mental health risks could reduce social media use among teenagers, with four states already passing legislation requiring such warnings. Separately, research presented at the European Respiratory Society conference suggested smoking contributes to increased loneliness, potentially due to cultural factors and reduced access to social activities.

These details were first reported by WDBJ7 in Roanoke, Virginia.