Medical Schools Lag on AI Training Despite Widespread Use

When Georgie Nahass organized programming lectures for fellow medical students at the University of Illinois Chicago, roughly 100 students showed up to the first session. The overwhelming response revealed a significant gap: while more than 80% of physicians currently use AI in their work, medical schools provide minimal education on these technologies.

A 2024 global survey spanning 48 countries found that 67.6% of medical, dental, and veterinary students held positive attitudes toward AI in healthcare. Yet less than 25% had received any formal education on it, and 75.3% reported little to no knowledge of AI. The disconnect between physician adoption and student preparation is creating anxiety among those entering the profession.

Why it matters

Medical students will practice in an AI-saturated healthcare environment without the training to use these tools safely or critically. This knowledge gap creates risks for patient safety, data privacy, and healthcare equity—particularly as students already use tools like ChatGPT during patient consultations without understanding their limitations or ethical implications.

Current Training Falls Short

Kailyn Geter, president of the Student National Medical Association at Howard University, sees attending physicians using AI but hasn't received formal training herself. Jeff Kim, an MD/PhD student at the University of Chicago Illinois, says his AI education consists of single slides tacked onto lecture ends or informal advice from practitioners.

A 2025 Association of American Medical Colleges survey found that while 77% of U.S. and Canadian medical schools offer AI coursework, only 39% provide training on data analysis and evaluation—skills experts consider more valuable than learning specific tools.

"There will always be a better tool within the next 2 years," said Dr. Felix Busch of the Technical University of Munich, who authored the 2024 global survey. He argues schools should teach how AI models are trained and where their data originates rather than focusing on individual applications.

Risks of Inadequate Preparation

Kim has observed peers consulting ChatGPT during patient encounters, excusing themselves to the restroom to identify diagnoses. This reliance prevents development of independent clinical reasoning skills.

Dr. Tina Nguyen, a bioethicist at The University of Texas Medical Branch, warns that students may be especially vulnerable to automation bias—the assumption that AI outputs can be trusted without human verification. "If they use the outputs and it's completely wrong, that blame is not going to be on the AI. It's going to be on the student," she said.

Privacy violations present another concern. Sarup Saroha, a medical student at University College London, notes that some clinicians face lawsuits from patients after using AI tools to record and transmit information without proper consent.

Barriers to Implementation

Despite calls from the AAMC and World Medical Association for AI training, institutions face obstacles including lack of faculty expertise, limited access to digital tools, and already-packed curricula. No standardized frameworks exist to guide what medical students should learn about AI.

Nahass compares the situation to MRI training: students don't need to know every technical detail, but they should understand what's happening inside the machine before using it. "With AI, though, those agreed upon tenets haven't really been established," he said.

Following his lecture series success, Nahass and University of Washington medical student Hugh Alessi founded Code Grand Rounds, a startup offering free AI and data science curricula to students and residency programs.

These details were first reported by Medscape.