Establishing Professional Standards for Responsible AI Adoption in Gastrointestinal Endoscopy

Article Impact Level: HIGH
Data Quality: STRONG
Summary of  Annals of Internal Medicine https://doi.org/10.7326/ANNALS-25-03415 
Dr. Omer F. Ahmad  et al.

Points

An international panel of fourteen experts from eleven countries developed a comprehensive framework to guide the responsible and ethical integration of artificial intelligence tools into modern gastrointestinal endoscopic medical practice.
The consensus reached through a rigorous two round Delphi process established ten essential statements focused on the critical domains of data governance and medicolegal accountability and healthcare equity and bias.
Experts emphasized that all artificial intelligence systems must maintain strict patient privacy through anonymization while ensuring complete transparency regarding algorithmic updates and institutional policies on clinical data ownership and usage.
The framework calls for professional societies to establish clear liability guidelines regarding adverse outcomes that may occur during artificial intelligence assisted diagnostic procedures or when using automated clinical report generation.
To prevent the widening of healthcare disparities researchers must ensure that training datasets are demographically diverse and that safety net hospitals gain equal access to these emerging diagnostic technology tools.

Summary

This international consensus statement, published in Annals of Internal Medicine, establishes the first comprehensive framework for the integration of artificial intelligence (AI) in gastrointestinal (GI) endoscopy. Led by experts from Beth Israel Deaconess Medical Center as part of the OperA project, the initiative addresses the legal and ethical barriers currently hindering clinical adoption. The framework provides a structured approach to navigating the complexities of data governance, medicolegal accountability, and health equity in an era of rapid technological advancement in colorectal cancer prevention.

The consensus was developed through a rigorous two-round Delphi process involving 14 international experts from 11 countries across four distinct global regions. Participants reached agreement on 10 key statements categorized into three primary thematic domains. In the domain of data governance, the panel emphasized the necessity for AI systems to comply with local protection regulations and maintain transparency during algorithmic updates. Medicolegal accountability remains a critical focus, with the experts calling for professional societies to provide specific liability guidance for AI-assisted diagnostic outcomes and automated reporting errors.

To mitigate bias and ensure equitable deployment, the framework mandates that AI systems be trained on demographically diverse datasets with transparent reporting of population characteristics. The experts highlighted the risk of widening healthcare disparities if advanced AI access is limited to well-resourced centers versus safety-net hospitals. As AI transitions from simple polyp detection to complex applications like automated report generation, this World Endoscopy Organization (WEO)-endorsed statement serves as a foundational guideline for regulators and clinicians to ensure that clinical innovation is matched by robust safety and ethical standards.

Link to the article: https://www.acpjournals.org/doi/10.7326/ANNALS-25-03415

References

Ahmad, O. F., Mori, Y., Bretthauer, M., Dourado, D. A., Hassan, C., Bisschops, R., Bhandari, P., Byrne, M. F., Dekker, E., Mahadevan, U., May, F. P., Messmann, H., Misawa, M., Ogata, H., Saito, Y., Silverman, A. L., Wang, P., Yano, T., Aabakken, L., & Berzin, T. M. (2025). The legal and ethical framework for artificial intelligence in gastrointestinal endoscopy: A world endoscopy organization international consensus statement. Annals of Internal Medicine, ANNALS-25-03415. https://doi.org/10.7326/ANNALS-25-03415