More Australians are increasingly turning to artificial intelligence (AI) chatbots for health-related information, prompting medical professionals to raise concerns about the limits and risks of relying on AI for medical advice. Pharmacists and general practitioners (GPs) report that patients frequently come to them seeking clarification or confirmation of diagnoses suggested by AI tools.
Pharmacy owner Laura Rheinberger, based in Mudgee, New South Wales, noted a growing trend of people, particularly parents of young children and individuals with sleep problems, consulting AI chatbots before visiting healthcare providers. Rheinberger said that while AI can provide preliminary information, it lacks the ability to perform essential diagnostic procedures such as physical examinations, blood tests, or imaging. “Sleep issues in particular can have different causes, from day-to-day stresses all the way to more complex issues like obstructive sleep apnoea,” she explained. Rheinberger also revealed that as a qualified pharmacist, she had personally consulted AI while attending to her newborn during the night, underscoring the technology’s growing integration into everyday health queries.
General practitioners have similarly observed an influx of patients presenting with self-diagnoses informed by AI. Dr. Sidd Sarmah, a Queensland-based GP, highlighted concerns about patients seeking quick and convenient answers, describing this phenomenon as a search for “sugar hits.” According to Dr. Sarmah, while AI operates through pattern recognition, human health conditions often exhibit variations that AI may overlook. “The human body isn’t a code to be understood or interpreted in a particular way,” he said.
Dr. Sarmah recounted a case involving a man in his 30s who believed, based on AI consultation, that his prolonged diarrhoea was a stomach bug contracted during overseas travel. Upon medical examination, the man was found to be anaemic and was ultimately diagnosed with Crohn’s disease—a condition that might have been overlooked if reliant solely on AI-generated information. Dr. Sarmah emphasized that while patient engagement with AI can encourage curiosity and self-education, bypassing professional medical evaluation carries risks, including delayed diagnosis of serious conditions.
Despite these cautionary views, no medical practitioners dismissed the potential for AI to support health education. However, they stress that AI should complement, not replace, in-person consultations and professional judgment.
The rapid adoption of AI in healthcare also aligns with broader discussions about its impact in other sectors, including education, where similar benefits and risks are being debated. As Australia navigates this technological shift, questions around data security, privacy, and the maintenance of professional standards are emerging alongside the promise of AI-enhanced learning and productivity.
In the context of healthcare, experts agree that while artificial intelligence can be a useful tool for providing general information and supporting patient engagement, comprehensive medical assessment and diagnosis remain reliant on human professionals capable of conducting hands-on evaluations and interpreting complex health variations.