HHS Identifies Three Priorities for Clinical AI Adoption

Federal health officials outlined three priorities Thursday that they said will accelerate artificial intelligence adoption in clinical care, focusing on better agency coordination, implementation guidance and clearer standards for evaluating AI tools.

Officials said those goals will shape future department actions as it implements AI. 

“We believe that starting with these three things and acting on constant engagement from this community is what’s needed to establish trust,” said HHS Deputy Chief AI Officer Arman Sharma. “And trust in this technology is the only thing that will lead to responsible but also effective adoption.” 

Thomas Keane, national coordinator for Health IT, said AI is already reducing administrative work, helping enroll patients in clinical trials, accelerating drug discovery and giving patients better access to health information.

“Most importantly, it’s empowering patients to understand and take control of their own health,” Keane said. 

7,000 Public Comments Shaped the Strategy

The Department of Health and Human Services identified the three priorities from the more than 7,000 comments submitted in response to a December request for information (RFI).

The RFI came as part of the Trump administration’s effort to accelerate clinical AI adoption across healthcare. Rather than focusing solely on AI development, the department sought recommendations on how to use its regulatory, reimbursement, and research and development authorities to speed implementation of AI technologies in clinical care. 

The department specifically requested feedback on modernizing digital health regulation, aligning reimbursement with AI-enabled care, supporting implementation research, reducing provider burden, improving patient outcomes and lowering healthcare costs. Officials said the RFI was intended to extend HHS’s internal AI strategy into a broader, department-wide effort to encourage AI adoption across the healthcare system. 

Adoption, Not Just Innovation 

A common theme in the public comments was the need for adoption support, according to Sharma. Building AI tools is only one part of the challenge, he said. Providers also need governance frameworks, implementation guidance and confidence that products have been adequately evaluated. 

“We recognize that even with the most agile regulation, even with optimal reimbursement, there’s still an implementation gap,” Sharma said. “There’s still process and workflow optimization that’s required to take AI and bring it into existing workflows within any care setting, whether that be a small clinic or whether that be a large academic medical center.” 

Sharma pointed to ARPA-H as an agency working to accelerate AI adoption in clinical care. Its Agentic AI-EnableD CardioVascular CAre TransfOrmation (ADVOCATE) program would be the first of its kind, FDA-approved clinical agentic AI system. Its goal is to “transform advanced cardiovascular disease management with an agentic AI system that can provide 24/7 holistic clinical care.”   

ARPA-H program manager and cardiologist Haider Warraich said the vision is for ADVOCATE to do everything a clinician can over the phone. 

“That could include administrative functions, such as helping set up appointments, helping provide advice about diet and lifestyle, which is so important for patients with cardiovascular disease, but also provide actual diagnoses, perform actual triage, and really perform high-risk functions, including being able to change medications or prescribe new medications,” he said. 

Warraich acknowledged concerns about introducing powerful technology and said he supported building a “supervisory layer” that would optimize human oversight of the technologies after they’ve been deployed in the postmarket. He also said implementing the program is “going to be incredibly challenging” and welcomed health systems to co-develop the technology and then deploy it in their clinical settings in randomized clinical trials. 

FDA Prepares Updated AI Policy 

Rick Abramson, director of the Food and Drug Administration’s Digital Health Center of Excellence, said the agency plans to release new policy ideas for stakeholder comment as it updates its regulatory approach to increasingly sophisticated AI-enabled medical devices. 

Abramson avoided discussing specific proposals because of ongoing policy development, but outlined several themes expected to guide future oversight. He said the FDA plans to address four areas that the public has consistently requested, including: 

• Greater clarity on what the FDA regulates, what is required of sponsors and what role FDA will play in both the premarket and post market settings. 

• Oversight proportional to risk. 

• Appropriate oversight across the lifecycle of AI products after they enter the market. 

• The need for policy coordination among federal agencies, between federal and state oversight bodies and also with international regulators. 

He noted that AI technologies evolve much faster than traditional regulatory frameworks, requiring the FDA to continuously adapt its approach as both the technology and clinician interaction with AI change. 

“The world is looking to the FDA for leadership in how to approach advanced clinical AI tools. We plan to meet the challenge, and the public should be expecting us to release some ideas to the public for stakeholder comment in very short order,” he said.