AI Morbidity and Mortality (AI M&M): A New Framework for Learning from Clinical AI Errors
Summarized by AI from reporting by ArXiv cs.AI, published under our editorial policy.
Researchers from Harvard Medical School and MIT propose AI Morbidity and Mortality (AI M&M), a structured framework to analyze individual AI-related errors and near-misses in healthcare, aiming to improve patient safety by understanding how risks emerge from interactions between AI systems, clinicians, and workflows.
Key takeaways
- AI Morbidity and Mortality (AI M&M) is a new framework proposed by researchers from Harvard Medical School and MIT to analyze individual AI-related errors and near-misses in healthcare.
- The framework includes incident reporting, root cause analysis, feedback loops, and continuous monitoring to enhance patient safety.
- AI M&M aims to bridge the gap left by current safety mechanisms, which are not designed to explain how risks emerge from complex interactions between AI systems, clinicians, workflows, and institutional controls.
Researchers from Harvard Medical School and MIT introduced AI Morbidity and Mortality (AI M&M), a new framework designed to improve the safety of clinical artificial intelligence. This structured approach aims to reconstruct and learn from individual AI-related errors and near-misses in healthcare settings.
Why Current Safety Mechanisms Fall Short for Clinical AI
Currently, clinical AI systems are monitored through aggregate model performance metrics and traditional patient safety reporting. However, these methods are not designed to explain how risks emerge from the complex interactions between AI systems, clinicians, workflows, and institutional controls. AI M&M is proposed as a solution to fill this gap by providing a structured, blame-free process for reviewing AI-related incidents.
The Four Components of the AI M&M Framework
The AI M&M framework involves several key components:
1. Incident Reporting: Clinicians and AI systems report errors and near-misses in a standardized format. 2. Root Cause Analysis: A multidisciplinary team reviews the incidents to identify underlying causes and contributing factors. 3. Feedback Loop: Insights from the analysis are used to improve AI systems, clinical workflows, and institutional policies. 4. Continuous Monitoring: The framework includes ongoing monitoring to ensure that lessons learned are applied and new risks are identified promptly.
Potential Impact on Patient Safety
AI M&M has the potential to significantly enhance patient safety by providing a systematic way to learn from AI-related incidents. This structured approach can help healthcare providers and AI developers identify patterns and implement improvements, ultimately reducing the risk of future errors. For patients, this means more reliable and safer AI-assisted care.
How to Get Started with AI M&M
If you are a healthcare professional or AI developer, you can start by familiarizing yourself with the AI M&M framework. Review the proposed guidelines and consider how they can be integrated into your practice or AI development process. For more detailed information, you can access the full paper on arXiv at https://arxiv.org/abs/2609.00076.
Frequently asked
- Who developed the AI M&M framework?
- The AI M&M framework was developed by researchers from Harvard Medical School and MIT.
- How does AI M&M differ from traditional patient safety reporting?
- AI M&M provides a structured, blame-free process for reviewing AI-related incidents, focusing on the interactions between AI systems, clinicians, and workflows, whereas traditional reporting is not designed to explain how risks emerge from these complex interactions.
- Where can I find more information about AI M&M?
- You can access the full paper on arXiv at https://arxiv.org/abs/2609.00076.