Much of the medical advice provided by free generative AI chatbots is inaccurate and have potentially dangerous limitations, reported Business Day (15 April 2026). According to a study published in the British Medical Journal (BMJ) almost half of the responses were found to be below acceptable standards; 20% were highly problematic and 30% were somewhat problematic. The research highlights concerns regarding the reliability of these chatbots as they cannot make ethical or value-based judgments and often produce authoritative-sounding but flawed responses.
The research (February 2026) focused on five free versions Gemini ChatGPT, DeepSeek, Meta AI and Grok.