Congress Pushes for Clarity on HHS' AI Assurance Lab Plans

Background and Concerns

With deregulation a priority for the incoming Trump Administration in 2025, Republicans have been closely monitoring how AI in healthcare will be regulated. In their letter to Micky Tripathi, who also serves as Assistant Secretary for Technology Policy Secretary and National Coordinator for Health IT, the representatives raised questions about the overarching objectives of the agency's reorganization and the potential for conflict of interest.

Part of a larger technology restructuring effort by HHS, the new ASTP - formerly Office of the National Coordinator for Health Information Technology - announced in July that it would have increased responsibilities, including over healthcare AI. However, the representatives believe that the creation of assurance labs could supplement the U.S. Food & Drug Administration's review of AI tools and lead to significant issues.

Specific Concerns about Assurance Labs

The representatives are particularly troubled by the possible creation of fee-based assurance labs that would be comprised of competing companies. They fear that larger, incumbent tech companies could gain an unfair competitive advantage and negatively impact innovation in the industry. This concern was highlighted when Miller-Meeks specifically asked if the FDA would outsource certification to the coalition, noting that Google and Microsoft are founding members while Mayo Clinic, which has more than 200 AI deployments, employs some of the coalition's leaders.

"It does not pass the smell test," she said, suggesting clear signs of an attempt at regulatory capture. CHAI, which unveiled standards for healthcare AI transparency, has said a long-awaited AI nutrition label will be coming soon. Dr. John Halamka, president of Mayo Clinic Platform, addressed the substantial potential benefits and real potential harms that could come from predictive and generative AI used in clinical settings earlier this year at HIMSS24.

"Mayo has an assurance lab, and we test commercial algorithms and self-developed algorithms," he said in March. "And what you do is you identify the bias and then you mitigate it. It can be mitigated by returning the algorithm to different kinds of data, or just an understanding that the algorithm can't be completely fair for all patients. You just have to be exceedingly careful where and how you use it."

Call for Responses and Ongoing Dialogue

The representatives included eleven questions and requested responses by December 20. A spokesperson for ASTP told Healthcare IT News by email that the agency is unable to comment on the letter at this time. CHAI has not responded to our request for comment, but this story will be updated if one is provided.

"The ongoing dialogue around AI in healthcare must consider the distinct authorities and duties of various agencies and offices to prevent overlapping responsibilities, which can lead to confusion among regulated entities," the four Republican members of Congress said in their letter. This highlights the importance of a clear and coordinated approach to AI regulation in healthcare.