AI Morbidity and Mortality: A Framework for Clinical AI Failure Review

Summary

The paper introduces AI Morbidity and Mortality (AI M&M), a structured, blameless process for reviewing clinical AI failures and near-misses. It complements model monitoring and patient-safety reporting by reconstructing how risk emerges across AI systems, clinicians, workflows, and institutional controls. The framework organizes each case around Trigger, Mechanism, Clinical Pathway, and Corrective Action. In five illustrative outpatient cases, two reviewers agreed on all 20 classification judgments. The authors call for prospective evaluation across institutions, systems, and care settings.