HHS launches month-long sprint to set clinical AI standards

Federal push for clinical AI standards

The White House and Department of Health and Human Services are launching a coordinated effort to establish standards for evaluating artificial intelligence tools used in clinical settings, according to an invitation to participate reviewed by STAT.

The initiative, organized by the White House Office of Science and Technology Policy, the Food and Drug Administration, and the Office of the National Coordinator for Health Information Technology, aims to produce "a consensus set of principles" for benchmarking and evaluating clinical AI systems.

How the sprint will work

The effort is structured as a one-month "sprint" divided into two phases: a written phase followed by a discussion phase. The compressed timeline suggests urgency among federal officials to address the rapidly expanding deployment of AI tools in healthcare without clear evaluation frameworks.

The invitation does not specify how many experts will participate or which organizations have been invited, but the multi-agency coordination signals a whole-of-government approach to the challenge of clinical AI oversight.

Why it matters

Healthcare AI systems are proliferating faster than the regulatory and evaluation infrastructure needed to assess their safety and effectiveness. Without standardized benchmarks, hospitals and clinicians lack consistent ways to compare competing AI products or verify vendor claims about performance. This sprint represents the federal government's most concrete step yet toward creating that infrastructure—potentially shaping how billions of dollars in AI investments are evaluated and how these tools are integrated into patient care. The involvement of both the FDA (which regulates medical devices) and ONC (which oversees health IT standards) suggests any resulting principles could have teeth across multiple regulatory domains.

Broader context

The initiative comes as healthcare organizations face mounting pressure to adopt AI technologies while lacking clear guidance on how to evaluate their clinical utility. Multiple federal agencies have issued preliminary frameworks and discussion papers on AI governance, but concrete standards for clinical validation have remained elusive.

The FDA has approved hundreds of AI-enabled medical devices, but critics have noted that many lack rigorous post-market surveillance or real-world performance data. Meanwhile, the ONC has been working to ensure health IT systems can safely incorporate AI capabilities without compromising interoperability or patient safety.

The one-month timeline is notably aggressive for developing consensus standards in a field as complex and contested as clinical AI evaluation. Whether such a compressed process can produce principles with broad stakeholder buy-in remains to be seen.

The details were first reported by STAT health tech correspondent Mario Aguilar, who reviewed the invitation to participate.