HHS Launches AI Clinical-Trial Program to Speed Studies With Fewer Participants

The U.S. Department of Health and Human Services is backing AI systems designed to make clinical trials faster, less costly and less burdensome for patients. On September 30, 2026, it launched SURPASS through the Advanced Research Projects Agency for Health, or ARPA-H, alongside three coordinated projects addressing trial sites, data access and patient navigation.

The effort targets a lengthy development process. HHS says developing drugs and biologics often takes more than a decade and fails more than 90% of the time. Its proposed response combines changes to how studies generate evidence with changes to the infrastructure needed to run them.

Simulate first, analyze continuously

SURPASS aims to build continuous, adaptive trials: studies that can change as evidence accumulates. The program would combine predictive computer models, shared trial infrastructure, common control groups and real-time analysis to help researchers make earlier decisions about treatments.

• Trial design: Integrate digital twins and other predictive models to simulate clinical and operational outcomes before a study starts. The goal is faster trials with fewer patients, while developing evidence that gives regulators confidence in these approaches.
• Continuous analysis: Build an engine that can analyze trial data in real time or on demand, support rapid adaptations and reduce the need for large conventional control groups while maintaining rigorous evidence.
• Automated operations: Create an agentic layer—software that carries out tasks—to automate trial startup and operations. It would help onboard new treatment arms and speed data collection, cleaning and the construction of datasets.

Sites, consent and a route into research

The three companion projects address barriers beyond study design. STACK aims to use AI to activate clinical sites faster, including helping facilities without research experience become capable of running trials. HHS says this would increase capacity and bring research opportunities closer to patients.

COMMONS is intended to connect nationwide data infrastructure, with privacy built into a system for managing consent at national scale. It seeks access to data suitable for regulatory use. CINCH focuses on patients contributing their own real-world data, coordinating care and finding potentially appropriate trials.

Up to $41.18 millionSTACK

Nextgov/FCW reports that Evidence Health is the prime awardee and could receive this amount.

Up to $49.9 millionCOMMONS

Nextgov/FCW identifies Evidence Health as the prime awardee, with this potential award.

Up to $8.95 millionCINCH

Nextgov/FCW identifies Courage Health as the prime awardee, with this potential award.

According to Nextgov/FCW, COMMONS targets more than 40 million individuals and 3 million linked records. The University of Texas at Austin Dell Medical School is partnering with Evidence Health on both STACK and COMMONS, preparing research sites and developing a national research data resource.

For organizations interested in SURPASS, a solution summary is due November 30, 2026. ARPA-H plans two informational sessions beforehand, according to Nextgov/FCW. The program's next step is soliciting approaches to build these systems, not deploying a finished trial platform.