INTRODUCTION

Artificial intelligence (AI) is expanding rapidly in healthcare, particularly in gastroenterology (GI). Applications include enhancing colonoscopy accuracy, identifying dysplasia, and predicting clinical outcomes in inflammatory bowel disease. Despite AI's potential, concerns about ethical use and inadequate training persist.

AIMS

This study aimed to evaluate the perceptions and understanding of AI among U.S. GI trainees and early career professionals.

METHODS

A cross-sectional survey was conducted among members of the American College of Gastroenterology. The survey assessed AI experience, confidence, and educational needs, with comparisons between fellows and early career. Fisher's exact test and multivariate regressions were applied to assess differences in responses.

RESULTS

A total of 193 respondents completed the survey (54.1% fellows, 45.9% early career). Most respondents (73%) were aged 25-35, and 91.6% believed AI would significantly impact healthcare. Only 6.7% currently use AI in practice, with fellows less likely to use AI compared to early career gastroenterologists (65.4 vs. 82.0%, P = 0.031). Confidence was low overall, with fellows reporting significantly less confidence in interpreting AI findings (P = 0.025) and understanding AI limitations (P = 0.050). A notable gap in the basic understanding of AI concepts, such as machine learning and neural networks, was evident among both groups, particularly fellows. Common concerns included misuse (64.2%), data privacy (40.2%), and financial costs (32.7%).

CONCLUSIONS

Interest in AI is high, but confidence and experience are limited. Ethical concerns and knowledge gaps highlight the urgent need for structured AI education in GI training. Comprehensive AI curricula focusing on practical use, critical interpretation, and ethics are strongly needed to equip gastroenterologists with the skills for responsible and effective clinical integration.