Background: Artificial intelligence (AI) has emerged as a transformative force in gastrointestinal (GI) endoscopy, aiming to enhance diagnostic accuracy, detection rates, and workflow efficiency. With multiple AI-assisted systems now reaching clinical use, there is a growing need to consolidate evidence regarding their performance, applications, and limitations. Objective: To synthesize contemporary evidence from randomized trials, meta-analyses, and real-world studies to provide a clinically oriented overview of the effectiveness, limitations, and implementation challenges of artificial intelligence in gastrointestinal endoscopy, and to identify key translational gaps that justify the need for this updated review. Methods: A structured narrative synthesis of literature from PubMed, Scopus, Web of Science, and manual reference screening (2005– 2025) was performed. Evidence was thematically analyzed across major domains including CADe, CADx, upper gastrointestinal neoplasia, capsule endoscopy, quality monitoring, implementation challenges, and real-world performance. Key lessons, translational barriers, and future research priorities were extracted. Quantitative performance estimates were derived from representative randomized controlled trials and meta-analyses and are presented as reported ranges rather than pooled analyses. Results: CADe systems have demonstrated a consistent 15– 20% relative increase in adenoma detection rate (ADR) compared to conventional colonoscopy. CADx algorithms achieve > 90% accuracy in differentiating neoplastic from non-neoplastic polyps, supporting “resect-and-discard” strategies. AI tools in upper GI endoscopy achieve diagnostic accuracies of 88– 96% for early esophageal and gastric neoplasia, outperforming non-expert endoscopists. Despite these benefits, barriers persist—dataset bias, lack of generalizability, medicolegal ambiguity, and regulatory inconsistency. Conclusion: AI has proven efficacy in improving detection and diagnostic precision in GI endoscopy. Future progress requires multicenter validation, standardized datasets, ethical frameworks, and clinician training to enable equitable, safe, and evidence-based integration into routine clinical practice. Keywords: artificial intelligence, gastrointestinal endoscopy, computer-aided detection, computer-aided diagnosis, adenoma detection rate, colonoscopy, Barrett’s esophagus
Yagnik et al. (Sun,) studied this question.