Over the past seven months, HHS has moved from talking about AI as a hospital efficiency tool to actively funding and defending the idea of AI that diagnoses, prescribes and treats patients largely on its own.
Here’s a look at how that shift has played out:
- On Dec. 4, 2025, HHS released a broad AI strategy covering five areas: governance, infrastructure, workforce, research and modernizing care. The plan asks the CDC, CMS, FDA and NIH to work together on AI.
- On Dec. 19, HHS asked the public for input on speeding up AI use in clinical care. Deputy Secretary Jim O’Neill called AI “a transformative force for good across America.” The request drew more than 7,000 public comments.
- On Feb. 23, the American Hospital Association responded to the HHS request for input. It asked HHS to use existing healthcare rules instead of writing new ones, wanted physicians involved in coverage denials and said Medicare doesn’t pay enough to cover the real cost of AI tools.
- Meanwhile, Utah put testing in practice. The state let AI startup Doctronic automatically approve prescription refills for patients with chronic conditions. Four months of results came out in May. The AI approved renewals 72% of the time, and doctors agreed with those decisions 91% of the time. The rest were sent to a doctor, though doctors thought that extra caution was unnecessary about a third of the time.
- Utah’s own medical board called for the pilot to be paused. Around the same time, Pennsylvania sued an AI company for letting a chatbot pretend to be a doctor.
- On June 4, a Washington Post story detailed the administration’s bigger plans: offering more than $50 million in grants for AI heart-care research and a path to license fully independent AI physicians. HHS adviser Amy Gleason said, “the genie is out of the bottle.” One doctor warned that misplaced trust in AI would eventually hurt someone.
- On June 8, hospital leaders shared mixed reactions. Most liked AI that expands access but disliked the idea of a true “AI physician.” One CIO called the term “absurd.” Others worried about liability and trust.
- On June 25, HHS held a webinar on what it learned from the public comments. The big reveal was ARPA-H’s ADVOCATE program, which aims to build the first FDA-approved AI agent for round-the-clock heart care. Program lead Haider Warraich, MD, said the goal is an AI that can “do everything a clinician can over the phone,” since nearly half of U.S. counties have no cardiologist. A second AI would supervise the first one. Nothing is ready for patients yet.
At the Becker's 11th Annual IT + Revenue Cycle Conference: The Future of AI & Digital Health, taking place September 14–17 in Chicago, healthcare executives and digital leaders from across the country will come together to explore how AI, interoperability, cybersecurity, and revenue cycle innovation are transforming care delivery, strengthening financial performance, and driving the next era of digital health. Apply for complimentary registration now.