Should You Trust AI Health Advice?

Artificial intelligence is increasingly being used to deliver health advice through chatbots, symptom checkers, and wellness apps, but experts warn that users should approach these tools with caution due to risks of inaccuracy, bias, and overreliance.

A recent BBC investigation highlighted growing concerns about the reliability of AI-generated health recommendations, particularly as more people turn to digital assistants for medical guidance instead of consulting healthcare professionals. The report noted that while AI can process vast amounts of medical data quickly, its outputs are not always clinically sound or tailored to individual patient needs.

AI systems can appear confident and authoritative, but they may lack the nuanced understanding required for safe medical advice, especially when dealing with complex symptoms or underlying conditions.

Dr. Alicia Chen, digital health researcher at Imperial College London

One of the primary risks identified is the potential for AI to generate plausible-sounding but incorrect or misleading information—a phenomenon known as “hallucination.” In health contexts, this could lead users to misinterpret symptoms, delay seeking care, or follow inappropriate self-treatment advice.

Studies have shown that some AI-powered symptom checkers perform poorly in identifying serious conditions. A 2023 review published in BMJ Health & Care Informatics analyzed 23 AI-driven tools and found that fewer than one-third provided accurate triage advice for emergency symptoms like chest pain or difficulty breathing.

Bias in training data also poses a significant concern. If an AI model is trained primarily on data from certain demographics—such as younger, urban, or higher-income populations—its recommendations may be less accurate or even harmful for underrepresented groups, including elderly patients, racial minorities, or those with rare conditions.

Regulatory oversight remains inconsistent across countries. While the U.S. Food and Drug Administration (FDA) regulates certain AI-based medical devices as software-as-a-medical-device (SaMD), many general wellness apps and chatbots fall outside formal oversight, meaning they are not required to undergo rigorous clinical validation before being released to the public.

The World Health Organization (WHO) has issued guidance urging caution in the use of AI for health, emphasizing that these tools should support—not replace—clinical judgment. In its 2021 ethics and governance framework for AI in health, the WHO stressed the importance of transparency, accountability, and human oversight in AI-assisted healthcare.

Experts recommend that users treat AI health advice as a starting point for information, not a definitive diagnosis. They advise verifying any concerning symptoms with a licensed healthcare provider and avoiding reliance on AI for managing chronic conditions, mental health concerns, or potential emergencies.

As AI continues to integrate into everyday health tools, developers and regulators face increasing pressure to improve accuracy, ensure equitable performance across populations, and establish clearer standards for safety and efficacy. Until then, public health officials urge skepticism and critical evaluation when seeking medical guidance from artificial intelligence.