Embedding artificial intelligence literacy in endoscopy training: extending the European Society of Gastrointestinal Endoscopy (ESGE) curriculum framework

10.1055/a-2742-4342

We commend the European Society of Gastrointestinal Endoscopy (ESGE) for their timely position statement providing a curriculum for the use of artificial intelligence (AI) [1]. The phased framework and emphasis on core endoscopic competency, AI literacy, and cognitive bias awareness provide an important foundation for safe implementation. We would like to offer a few additional suggestions that may strengthen future iterations of this curriculum.

First, the optimal point for incorporating AI into training remains uncertain. Early exposure may reinforce careful inspection habits; a recent cohort study reported higher long-term detection performance among endoscopists trained with computer-aided detection (CADe) compared with conventional training [2]; however, premature use risks dependency, cognitive overload, and automation bias. Trainee surveys similarly reflect enthusiasm tempered by uncertainty regarding the appropriate timing [3]. Clarifying when AI supports rather than substitutes skill acquisition should be a priority for prospective studies.

The discussion of automation bias and algorithm aversion is welcome. The evolution of scope-guide use illustrates how clinical tools reshape, rather than erode, skill sets: trainees learn to integrate new visual information alongside tactile and luminal cues, and can adapt when the tool is absent. AI may follow a similar trajectory. Rather than withholding AI during training, curricula could focus on defining explicit competencies for interacting with AI, including responding to uncertain predictions and avoiding over- or under-trust. Simulation platforms creating discordant human–AI scenarios may offer an effective way to expose and correct these biases [4].

Furthermore, while adenoma detection rate remains important, additional metrics, such as false-positive burden, workflow impact, interval lesion rates, and cost-effectiveness, would better reflect real-world AI performance.

In conclusion, we again congratulate the authors for addressing an emerging and complex area of clinical training. Further refinement in these areas may help ensure that the curriculum evolves into a comprehensive, competency-based framework that guides the safe and equitable integration of AI across diverse endoscopic practices.

Publication History

Article published online:
20 March 2026

• References

• 1 Mori Y, Kopylov U, Sinonquel P. et al. Curriculum for safe and effective use of artificial intelligence in endoscopy: European Society of Gastrointestinal Endoscopy (ESGE) Position Statement. Endoscopy 2025;
• 2 Orzeszko Z, Gach T, Necka S. et al. The implementation of computer-aided detection in an initial endoscopy training improves the quality measures of trainees' future colonoscopies: a retrospective cohort study. Surg Endosc 2025; 39: 5276-5286
• 3 Magahis PT, Pence CJ, Wan D. S758. Impact of Artificial Intelligence on Gastroenterology Training and Education: A Survey of Fellows' Perspectives. Am J Gastroenterol 2025; 118: S555-S556
• 4 Clement David-Olawade A, Aderinto N, Egbon E. et al. Enhancing endoscopic precision: the role of artificial intelligence in modern gastroenterology. J Gastrointest Surg 2025; 29: 102195