Artificial intelligence (AI) is increasingly being used in medical education, with new research showing its potential to support training, knowledge access and clinical learning, while reinforcing the continued importance of human expertise in patient care.
Several advanced AI systems are being explored in medical training, including Google’s Med-PaLM 2, OpenAI’s GPT-4 and specialised biomedical models such as BioBERT and ClinicalBERT. These tools are trained on large medical datasets to better understand clinical language, research literature and patient-care scenarios.
Studies suggest that some models are capable of performing at a high level in structured assessments. Med-PaLM 2 has demonstrated around 86.5% accuracy on MedQA benchmark questions, while GPT-4 has met passing thresholds on USMLE-style medical exam simulations, indicating strong knowledge recall in exam settings.
Expanding Role in Medical Learning
Educators and researchers are exploring how AI can support medical training in several ways:
-
Simulating patient - doctor interactions for learning
-
Generating study materials and assessment questions
-
Summarising complex medical research
-
Supporting clinical reasoning exercises
These applications position AI as an educational and decision-support tool rather than a clinical authority.
Important Limitations
Despite their capabilities, AI systems still face key limitations. Researchers note that they can generate inaccurate information, lack real-world clinical judgment and empathy and remain less reliable than trained specialists in interpreting complex medical imaging or making nuanced treatment decisions.
As a result, experts emphasise that AI tools should be used to complement medical education and clinical workflows, not replace healthcare professionals.
Why This Matters
The growing role of AI in healthcare education reflects broader changes in how medical knowledge is accessed and applied. While these technologies may enhance training and support learning, safe and effective care continues to depend on clinical expertise, ethical judgment and human interaction.
Sources:
https://pmc.ncbi.nlm.nih.gov/articles/PMC10948644/
https://www.nature.com/articles/s41591-024-03423-7
Reviewed by: Dr Cheryl Yeo
Bio: Dr Cheryl Yeo is the founder of iAM Health. She has a PhD in food science and nutrition, dedicating more than a decade to research on metabolic disorders and weight management, functional food ingredients, and the science behind metabolism-focused supplements and microbiome-driven health therapies. Her work deepened her understanding of how targeted nutrition, including probiotics and digestive support, can influence long-term wellbeing.
Connect with Dr Cheryl Yeo: cheryl@iamhealth.sg