Australians increasingly turn to AI for medical advice, alarming healthcare professionals

At a pharmacy counter in Mudgee, a regional town in New South Wales, something has shifted. Customers walk in with questions that would have seemed unusual just a year or two ago: Is this diagnosis from an AI chatbot real? Should I trust what the algorithm told me about my child's rash, or my own sleeplessness?

Laura Rheinberger, who owns Blooms The Chemist, has watched this pattern emerge among her regular customers. The inquiries cluster around two groups: parents worried about their young children, and adults struggling with sleep. People arrive having already consulted an AI system about their symptoms, seeking validation or clarification from someone with actual medical training. Rheinberger herself has done it—late at night with a crying newborn, she has turned to a chatbot for guidance on how to soothe her child. She understands the impulse. But she also understands its limits. An AI cannot examine a patient, cannot listen to a heartbeat, cannot order blood work or imaging. It cannot feel the texture of a symptom the way a trained hand can. Sleep problems alone illustrate the problem: they can stem from daily stress or from something far more complex, like obstructive sleep apnoea. Without a healthcare professional walking through the specifics of a person's life and body, the algorithm offers only pattern-matching—and patterns in medicine are not always what they seem.

General practitioners across Australia are encountering the same phenomenon. Sidd Sarmah, a GP in Queensland, describes it as a hunger for convenience, what he calls "sugar hits" in everyday life. People want answers fast, and they want them without leaving home. Sometimes this impulse helps; sometimes it creates larger problems downstream. Sarmah has noticed more patients arriving at his practice already convinced they understand their condition, armed with information pulled from an algorithm. He does not discourage people from learning about their own health—questioning what happens in one's body is important. But there are real dangers in bypassing a doctor's judgment entirely.

Consider a man in his 30s who came to Sarmah's practice complaining of persistent diarrhoea. The man had travelled extensively for work, and an AI system had suggested he had picked up a stomach bug abroad. He had been taking over-the-counter anti-diarrhoea medication, assuming the problem would resolve on its own. He had lost weight, which he attributed to not eating properly during his illness. When Sarmah reviewed his medical history, nothing suggested bowel problems. Basic tests revealed anaemia. The diagnosis, once properly investigated, was Crohn's disease—a chronic inflammatory condition that requires ongoing management and treatment. An algorithm, pattern-matching against common causes of diarrhoea, would have confidently pointed toward gastroenteritis. A doctor, looking at the full picture, saw something else entirely.

This is the gap that concerns Sarmah and his colleagues. Artificial intelligence operates through pattern recognition, which works well when cases fit neatly into established categories. But the human body resists neat categorization. A lump in the neck might be a swollen gland from a viral infection, or it might be something far more serious. Without physical examination, without the ability to ask follow-up questions grounded in what a doctor observes, an algorithm cannot distinguish between the two. It cannot know what it does not know.

Sarmah calls this moment "uncharted territory." The technology is evolving faster than medicine can fully understand its implications. But he is clear about one thing: general practitioners will remain central to healthcare, whatever tools patients consult along the way. A doctor's role is not simply to provide information—it is to interpret information in the context of a specific person's life, history, and body. That interpretation requires something an algorithm cannot offer: the ability to see, listen, touch, and feel a patient, and to adjust understanding based on what those senses reveal. As AI becomes more embedded in how people think about their health, that distinction may matter more than ever.