As the adoption of augmented intelligence (AI) tools in health care accelerates, physicians face the critical task of understanding and managing the risks associated with these technologies. Recognizing that AI is intended to support—not replace—clinical decision-making, the AMA AI Specialty Collaborative has developed this resource to provide physicians with a structured, practical framework for evaluating AI applications in medicine.
This guide emphasizes clinical relevance, transparency and patient safety while remaining accessible to clinicians without specialized technical expertise. The guide introduces five core domains for AI evaluation:
- Clinical Use Case and User
- Training and Validation Data Relevance
- Risks and Mitigation
- Effectiveness and Performance
- Workflow Integration and Monitoring
Cross-cutting considerations—including representation, transparency, clinical relevance and ongoing safety—help ensure a comprehensive assessment.
Download the guide (PDF).
