Clinical artificial intelligence is becoming embedded in real‑world care, yet current safety mechanisms struggle to reconstruct and learn from individual AI‑related errors or near‑misses. Aggregate model monitoring can spot performance shifts, and traditional patient safety reporting can capture adverse events, but neither explains how risk emerges from the interaction of AI systems, clinicians, workflows, and institutional controls. We propose the AI Morbidity and Mortality (AI M&M) framework, a structured, blameless case‑review approach focused on clinical AI failures. The framework combines standardized case intake, evidence preservation, investigator‑level reconstruction, tool‑in‑loop attribution, and corrective‑action tracking. Each event is classified across four linked dimensions: trigger, mechanism, clinical pathway, corrective action. This separates the condition that exposed a vulnerability, the process that generated risk, its impact on care, and the remediation assigned. We demonstrate the framework with five outpatient medication and clinical decision‑support cases; two clinician reviewers independently applied all four classification axes and achieved agreement on all twenty axis‑level labels. AI M&M is intended to complement, not replace, model monitoring, patient safety reporting, and regulatory oversight, converting individual AI‑in‑workflow failures into actionable institutional learning. Prospective evaluation across institutions, AI systems, and clinical settings is needed.
Review: The framework offers a systematic, blame‑free method for reviewing clinical AI failures, enabling organizations to quickly pinpoint risk sources and implement corrective measures.