Pragmatic Approaches to the Evaluation and Monitoring of Artificial Intelligence in Healthcare

Published: November 10, 2025

Figure1. The AHA has a proven track record of generating high quality scientific evidence, developing clinical guidelines and engaging education materials that are broadly disseminated, and accelerating the translation of this knowledge into action through the vast network of hospitals participating in its quality improvement registries and collaboratives. This operating model positions the AHA to be a trusted leader in developing, testing, and implementing processes to support the responsible use of AI in health care.
  • AI adoption in health care is accelerating faster than traditional evaluation frameworks, creating urgent challenges around safety, fairness, and clinical impact.
  • Effective governance requires a structured, risk-proportionate approach across pre-deployment, implementation, and post-deployment phases, anchored in strategic, ethical, clinical, and financial principles.
  • Sustainable and responsible AI integration depends on continuous real-world evidence generation, transparent monitoring, and collaboration among health systems, professional societies, and clinicians to safeguard patient outcomes.