AI Usage Among Medical Learners
Based on the most recent high-quality data, approximately 40-60% of medical learners have no prior experience with AI tools, meaning roughly 40-60% are using or have used AI in their medical education.
Current Usage Patterns
The most recent cross-sectional study from 2024-2025 provides the clearest data on actual AI usage among medical learners 1:
- 59.9% of medical students, residents, and faculty reported having NO prior experience with AI tools, which inversely suggests that approximately 40% have some experience using AI 1
- Despite this limited hands-on experience, 86.95% understood AI concepts at a basic level 1
- Only 12.04% reported being "very familiar" with AI applications in medical education, indicating that while some learners may be using AI tools, deep engagement remains uncommon 1
The Familiarity-Usage Gap
There is a striking disconnect between interest and actual implementation 1:
- 87% of medical learners support AI integration into medical curricula, yet actual usage remains below 50% 1
- 91.3% believe AI could improve educational efficiency, but this enthusiasm has not translated into widespread adoption 1
- 88.62% expressed desire for structured AI training, highlighting that lack of formal education may be limiting current usage rates 1
Important Caveats
The available evidence does not provide precise percentage data on current AI usage rates. The studies focus more on perceptions, applications, and potential rather than quantifying actual user adoption 2, 3, 4. The 40-60% estimate is derived from inverse reporting of non-users rather than direct measurement of active users 1.
The data comes primarily from a single institution in India, which may not represent global patterns 1. Multi-center studies are needed to validate these findings across different educational systems and geographic regions 1.
Clinical Context
The low usage rates occur despite AI tools being increasingly available for medical education applications including exam question generation, clinical reasoning development, diagnostic skill improvement, and surgical training assessment 3, 4. This suggests that barriers to adoption—such as lack of structured training, unclear integration pathways, and limited faculty expertise—are more significant obstacles than tool availability 1.