Abstract
National artificial intelligence (AI) policy in the United States is accelerating rapidly through initiatives such as America's AI Action Plan, creating urgency for health care organizations to respond. Although these policies are not health care–specific, their implementation has immediate implications for clinical practice because AI is increasingly operationalized through electronic health record (EHR)–embedded workflows. This article argues that a critical policy and practice gap exists: National AI strategies emphasize speed and innovation while insufficiently addressing health care's core operational reality—the EHR as AI infrastructure—and the preparedness of the clinical workforce to safely use AI tools. Using a conceptual framework that links national AI policy momentum, EHR infrastructure, and workforce education, this article demonstrates that most health care organizations already deploy production-level AI through EHRs, including ambient documentation, clinical decision support, predictive analytics, chart summarization, documentation integrity tools, and patient-facing AI portals. As these tools become routine, clinicians increasingly function as the “human in the loop,” bearing responsibility for AI-informed decisions without consistent training or governance support. The article highlights evidence from nursing education and workforce studies demonstrating gaps in EHR and AI readiness, particularly among students and newly graduated clinicians. It concludes that workforce education constitutes essential “human infrastructure” for health care AI. Health care executives—especially those accountable for quality and safety—must actively engage in shaping AI implementation by aligning policy advocacy, EHR governance, and EHR-based education to ensure that accelerated AI adoption improves care rather than amplifies risk.