A study of artificial intelligence literacy among infectious diseases faculty and fellows at an academic medical center identified differences in clinical exposure, preparedness and learning needs, highlighting the need for structured AI training in fellowship programs.
The study, presented at IDWeek 2026, found that nearly three-quarters of faculty reported regular or occasional AI use, but rated their preparedness to use the technology responsibly as moderate (mean score, 3.2 out of 5). Among fellows, five of six (83%) reported limited exposure to AI in clinical practice.
The prospective, cross-sectional study involved 19 faculty members and six fellows in an academic infectious diseases program. Separate surveys assessed AI use, clinical applications, attitudes toward integration, perceived competency, knowledge gaps and preferred learning formats.
Faculty commonly used AI to review literature, summarize clinical guidelines and support differential diagnoses, the authors reported. However, gaps remained in their understanding of AI limitations and ability to assess potential bias and accuracy.
All faculty respondents agreed or strongly agreed that AI poses risks without oversight. Most also viewed AI literacy as essential for future infectious diseases physicians. Despite their AI experience, faculty remained uncertain about its responsible use.
Fellows most often cited limited access to technology and accuracy concerns as barriers to AI use. Insufficient training and privacy concerns were additional obstacles, the authors found.
According to the study, the fellows wanted to learn about AI-driven diagnostic tools, antimicrobial stewardship applications and ethical considerations. They preferred case-based, hands-on learning, whereas faculty favored informal discussions and structured teaching sessions.
“There is a clear readiness gap at multiple levels: faculty use AI clinically but lack confidence in responsible implementation; fellows have minimal structured exposure despite high interest in learning,” the authors wrote.
The study authors also emphasized faculty concerns about using AI without oversight and fellows' preferences for practical learning opportunities. “Universal faculty concern about unsupervised AI use, paired with fellows’ preference for case-based formats, provides a clear mandate for competency-oriented AI education in ID fellowship training,” the authors wrote.
They noted that the findings may help guide the development of a structured AI curriculum designed for the infectious diseases fellowship learning environment.