Roughly one in six people are now turning to artificial intelligence tools for health advice, according to The Times, a figure the paper reports has prompted a warning about medical misinformation.
That headline number is striking mostly for how ordinary it makes the behavior sound. Asking a chatbot about a symptom, a medication, or a worrying test result is no longer a fringe habit. It is something a sizable slice of the public does routinely, largely outside the view of the doctors who would otherwise field those questions.
The warning accompanying the figure points at the core problem with the practice: AI systems can produce confident, fluent answers that are wrong. In most contexts a wrong answer is an inconvenience. In health, it can mean someone delays care they need, or seeks care they don't.
The view from inside medicine is not uniformly alarmed. In an interview published by calcalistech.com, the argument was framed simply: "AI tools will not replace doctors, but they will save us work." That captures a distinction worth holding onto. AI handling paperwork, summaries, and routine administrative load inside a clinical setting is a different proposition from a patient consulting a chatbot alone at home, with no clinician checking the output.
The two framings are already colliding. Adoption by patients is running ahead of any settled guidance on what these tools should and shouldn't be trusted to answer, and ahead of the professional consensus about where they belong in care.
This matters because health is the area where the gap between an AI that sounds authoritative and one that is actually correct carries the highest cost — and a sixth of the population is already living inside that gap.