Thoughts on Healthcare Markets & Technology

Thoughts on Healthcare Markets & Technology

HHS Is Funding Agentic AI in Clinical Care With $1.5M in LEAP Awards to MedStar and Regenstrief. The Bigger Challenge Is Who Controls the Agent: Patient, EHR Vendor, Hospital, Payer, AI Lab

Oct 12, 2026
∙ Paid

🎧 Podcast episode for paid subscribers only. Also available on Apple Podcasts and Spotify.

To listen to paid episodes in Apple or Spotify, link your Substack subscription via the show settings on those platforms (instructions inside the Substack app under Subscriptions → Podcast).

Table of Contents

Nine seconds of health spending

What the notice asked for, and who was allowed to answer

The Sovereign Execution Bridge, translated out of grant-speak

Identity and authorization, or who is this thing and who said it could

Audit and provenance, or two fields and a prayer

Clinical responsibility and the liability hot potato

Five candidates for the keys

The lab award is the boring half that matters

What to build, what to watch

Thoughts on Healthcare Markets & Technology is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

Abstract

ONC announced the 2026 LEAP in Health IT awardees on Oct 6: two cooperative agreements totaling approximately $1.5 million.

Area 1 (agentic AI, up to $1 million) went to MedStar Health Research Institute for an open-source middleware layer called the Sovereign Execution Bridge.

Area 3 (laboratory interoperability, up to $500,000) went to Regenstrief Institute for a toolkit to assess readiness, identify terminology and interoperability gaps, and guide remediation at small independent laboratories.

Area 2 (a community feedback layer for FHIR endpoints through Lantern, ONC’s FHIR endpoint monitoring tool, with a $500,000 ceiling) received no award; ONC said it received applications only for Areas 1 and 3.

MedStar’s design: durable long-running workflows, patient-gated Solid pods, and cryptographically verifiable AI routing with tamper-evident certificates.

The dollars are small. The funded design is the news: the patient holds the keys, while the commercial market is handing them to EHR vendors, health systems, payers and chatbot companies.

Five things an agent needs that a chatbot never did: identity, authorization, audit, provenance, clinical responsibility.

Backdrop: HTI-5 still pending, FDA’s generative AI paper, ARPA-H’s ADVOCATE, Epic’s Agent Factory, America.gov, and a summer of test agents wandering where nobody invited them.

Company shapes: execution brokers, receipts and audit, agent credentialing, clinical saga libraries, lab code remediation.

User's avatar

Continue reading this post for free, courtesy of Thoughts on Healthcare.

Or purchase a paid subscription.
© 2026 Healthcare Markets & Technology · Privacy ∙ Terms ∙ Collection notice
Start your SubstackGet the app
Substack is the home for great culture