During his confirmation hearings, health secretary Robert F. Kennedy Jr. called on technology to solve America’s rural health care crisis. “President Trump is determined to end the hemorrhage of rural hospitals, and he’s asked me to do that through use of AI, through telemedicine,” Kennedy told senators, invoking the example of an AI nurse “that has diagnostics as good as any doctor.”
Kennedy’s appeal to technology coincides with a massive push to adopt AI across the American health care system, but his vision is far from reality. AI algorithms built on patient data from one location — typically a well-resourced, coastal health system — often struggle to perform similarly when they’re used elsewhere, like a rural Wyoming hospital. Narrowing those performance gaps has been a central goal of clinical AI developers.
But some of that work is being delayed and stifled in response to President Trump’s early executive actions. Researchers, health systems, and regulators working to make AI safe and effective for diverse populations are reeling as they struggle to interpret the impact of executive orders, communications freezes, and staff reductions across federal health agencies. As researchers prepare for deeper cuts, the uncertainty is already having a chilling effect on clinical AI research.
This article is exclusive to STAT+ subscribers
Unlock this article — and get additional analysis of the technologies disrupting health care — by subscribing to STAT+.
Already have an account? Log in
