UPDATE: The American Heart Association (AHA) has just released urgent new guidelines for the responsible use of artificial intelligence (AI) in health care, addressing critical gaps in the evaluation and monitoring of AI tools. With hundreds of AI health tools approved by the U.S. Food and Drug Administration (FDA), only a small fraction have undergone rigorous assessment for their clinical impact, fairness, and potential biases.
Published today in the AHA’s flagship journal, Circulation, the advisory titled “Pragmatic Approaches to the Evaluation and Monitoring of Artificial Intelligence in Health Care” introduces a vital, risk-based framework aimed at enhancing the safety and effectiveness of AI in cardiovascular and stroke care. This advisory is essential as AI rapidly transforms health care, often outpacing traditional evaluation methods.
Dr. Sneha S. Jain, M.D., M.B.A., a key contributor to the advisory, stated, “AI is transforming health care faster than traditional evaluation frameworks can keep up. Our goal is to help health systems adopt AI responsibly.” The AHA emphasizes that effective AI governance is crucial for selecting, validating, implementing, and overseeing AI tools in patient care.
The advisory outlines four guiding principles for health systems to follow: strategic alignment, ethical evaluation, usefulness and effectiveness, and financial performance. These principles are designed to ensure that AI tools not only deliver measurable clinical benefits but also protect individuals from potential harms.
A recent survey revealed alarming statistics: only 61% of hospitals utilizing predictive AI tools validated them on local data before deployment, and fewer than half assessed for bias. This inconsistency raises serious concerns regarding equitable care delivery, particularly among smaller and rural institutions.
Highlighting its commitment to responsible AI governance, the AHA has pledged over $12 million in research funding for 2025 to explore novel AI delivery strategies that prioritize safety and efficacy. The AHA’s extensive network, which includes nearly 3,000 hospitals and over 500 rural and critical access facilities, positions it as a trusted authority in advancing these critical guidelines.
Dr. Lee H. Schwamm, M.D., FAHA, emphasized the importance of ongoing monitoring, stating, “Responsible AI use is not optional, it’s essential. This guidance provides practical steps for health systems to evaluate and monitor AI tools, ensuring they improve patient outcomes and support equitable, high-quality care.”
As AI technology continues to evolve, the AHA stresses that monitoring must not cease post-deployment. AI tools may drift in performance as clinical practices change or patient populations evolve, necessitating integration of AI governance into existing quality assurance programs.
The AHA’s new advisory not only addresses immediate challenges but also sets the stage for a safer, more equitable health care landscape. As health systems globally begin to implement these guidelines, the impact on patient care will be closely watched.
For more information, refer to the comprehensive advisory by Dr. Sneha S. Jain et al. in Circulation, DOI: 10.1161/cir.0000000000001400. As health care continues to embrace AI, these steps are critical to ensuring technology enhances care rather than complicates it.
