In a viewpoint letter published in JAMA on 31 March 2025, Marc D. Succi, Bernard S. Chang and Arya S. Rao address an urgent and timely question in modern medical education: how can we best prepare future clinicians to thrive in an era dominated by artificial intelligence (AI)? As AI continues to revolutionise various sectors, its impact on healthcare, particularly in medical training, is becoming increasingly profound.
The viewpoint, titled “Building the AI-Enabled Medical School of the Future,” explores the promise and challenges of incorporating large language models (LLMs) into medical education. These AI-powered systems have demonstrated early successes in clinical tasks, performing at levels comparable to or even surpassing human clinicians in certain settings. Yet, the authors caution that AI’s role in medical education must go beyond simply replicating human decision-making processes. Instead, it should aim to enhance, rather than replace, the critical thinking and clinical reasoning that define excellent medical practice.
AI in the Classroom: A Tool for Clinical Reasoning
The authors highlight the potential for LLMs to reshape how medical students are taught. Traditional medical education emphasises clinical reasoning—teaching students not just facts, but how to think through complex, dynamic clinical scenarios. However, current LLMs, while effective at recognising patterns and generating plausible answers, do not engage in true logical reasoning. This discrepancy poses a challenge for educators seeking to incorporate AI as a teaching tool.
The research suggests that LLMs could serve as powerful partners in medical education, particularly for case-based learning. These models can act as tutors, offering instant feedback on students’ clinical reasoning, providing critical evaluations of their logic, and facilitating interactive learning through simulated patient encounters. In a future medical school, LLMs could offer students exposure to a broader range of clinical cases, including rare diseases and culturally diverse patient presentations, which they might not encounter during traditional rotations.
Addressing the Gaps: Standardising and Democratising Education
One of the most promising aspects of AI integration into medical education is its potential to level the playing field across institutions. In resource-constrained settings, where patient diversity or volume may be limited, LLMs could provide students with a comprehensive set of clinical experiences. These AI-driven tools could also standardise the educational experience, ensuring that all medical students, regardless of their geographic or institutional background, have access to a robust and varied curriculum.
However, the integration of these technologies must be accompanied by an understanding of their limitations. The authors stress that while LLMs are valuable for expanding educational access and ensuring scalability, they must be used in conjunction with proven educational methods to preserve clinical rigor and human elements of care. Medical education must continue to nurture physicians who possess high-level skills in clinical reasoning, data interpretation, and ethical decision-making.
Preparing Physicians for AI-Augmented Practice
Looking ahead, the authors argue that medical education must evolve to teach students not only how to use AI tools effectively but also how to critically assess and integrate AI-generated outputs into patient care. Clinicians of the future will need to be experts in data systems, able to navigate AI-generated recommendations while ensuring that the technology aligns with clinical standards and ethical guidelines.
Importantly, this technological sophistication should not come at the expense of core clinical skills. The research underscores the importance of maintaining proficiency in traditional practices, such as history taking, physical examination, and differential diagnosis, so that clinicians are well-equipped to provide exceptional care even in situations where AI tools may fail or be unavailable.
Conclusion: A Collaborative Future for Medicine
The authors conclude by emphasising that the medical school of the future must thoughtfully integrate AI technologies, using them to enhance both education and patient care while upholding the principles of rigorous reasoning, empathy, and moral judgement. As AI continues to develop, collaboration between medical educators and machine learning experts will be critical to ensure that these tools are harnessed responsibly.
In sum, while the future of medical education will undoubtedly be shaped by AI, its true value will lie in its ability to complement and augment human expertise, creating a new generation of physicians who are not only technologically proficient but also deeply attuned to the ethical and humanistic aspects of care.
Paper: Succi MDChang BSRao AS. Building the AI-Enabled Medical School of the Future. JAMA. Published online March 31, 2025. doi:10.1001/jama.2025.2789
