Artificial Intelligence in Digestive Endoscopy Training-The Past, Present, and Future.

BACKGROUND AND OBJECTIVE: Artificial intelligence (AI) is reshaping gastrointestinal endoscopy, yet its role in training remains unexplored. This narrative review summarizes current evidence on AI-assisted endoscopy training, addresses potential drawbacks, and envisions future directions. METHODS: This narrative review was performed via a systematic MEDLINE search (including articles from inception to January 2025), with search terms covering 'AI', 'endoscopy,' and 'training.' Studies were excluded if they were reviews, letters, editorials or comments; focused solely on model development; lacked a training component; or were limited to simple comparisons between the performance of endoscopists and AI systems. After screening 1443 records, 27 articles were included in this review. RESULTS: AI demonstrates potential in enhancing the training of various types of endoscopy (including luminal, hepatobiliary, capsule, and therapeutic endoscopy) by improving quality metrics, enhancing lesion detection, and guiding anatomical landmark recognition, yet the current applications are mainly task-based. Future AI must evolve to provide comprehensive training and personalized performance tracking to endoscopists of different levels of experience. Further studies are needed to assess objective educational outcomes and cost-effectiveness. Key concerns for AI adoption, including deskilling, over-reliance, ethical considerations, and practicality, should be addressed through structured implementation, quality assurance, and regulatory framework. CONCLUSION: In conclusion, AI can augment endoscopy training by improving skill acquisition and procedural quality, yet significant gaps remain. More research is needed to support its widespread integration.