The U.S. Department of Health and Human Services announced Sept. 30 a federal program that aims to use artificial intelligence, computer modeling, and real-time data to test new drugs "faster, at lower cost, and with fewer participants." The announcement did not include a dollar figure for the effort.
The program, called SURPASS, will run through the Advanced Research Projects Agency for Health, or ARPA-H. A companion project would build national-scale data connections and a consent system for patient health records. HHS described that system as "privacy-by-design" but did not explain how patient data would be protected, who could access it, or how it would meet federal privacy law.
For patients, the stakes are concrete. Faster trials could bring treatments to market sooner. But the plan also relies on real-world health data from ordinary Americans, which makes privacy and oversight details more than a technical footnote.
A Plan to Rebuild Clinical Trials With AI
According to the HHS announcement, SURPASS stands for Simulation-augmented, Real-time Platform Adaptive Seamless Trials. It has three parts: a design engine that uses digital twins and other predictive models, a system that analyzes trial data as it comes in, and an AI layer to automate trial startup and operations.
HHS also announced three companion projects. STACK would use AI to bring more U.S. trial sites online, including sites new to research. COMMONS would connect health data nationally and create a consent system "at national scale." CINCH would help patients contribute their own real-world data and find trials faster.
HHS cited familiar problems. Developing a drug often takes more than a decade, costs up to $2 billion on average, and fails more than 90% of the time.
"America should lead the world in turning medical breakthroughs into treatments for patients," said HHS Secretary Robert F. Kennedy Jr. Daria Fedyukina, the SURPASS program manager, said today's system "often requires too many stops, too much duplicated infrastructure, and too much time before researchers can understand whether a trial is on the right track."
Funding and Privacy Details Remain Unstated
The HHS release contains no budget figure. STAT reported that "the initial announcement did not note how much funding was designated for the new initiatives" and described SURPASS as a five-year program. Axios reported that winning teams would receive funding and support over five years, without naming an amount.
A schedule for applicants does exist. The ARPA-H program page lists a webinar on Oct. 15, a proposers' day on Nov. 6, solution summaries due Nov. 30, and full proposals due Jan. 22, 2027. HHS has not said when results might reach patients.
On privacy, the release does not mention HIPAA, data security standards, or limits on how patient records may be used. Under the HIPAA Privacy Rule, researchers generally need patient authorization, de-identified data, a limited data set with a data use agreement, or a waiver from an ethics review board to use health information. HHS has not said which approach COMMONS or CINCH would follow.
The FDA Question
Any faster trial method helps patients only if regulators accept the results. The HHS release refers to "regulatory confidence" but does not name the FDA. Axios reported that the FDA was included in planning the program and that FDA Acting Commissioner Kyle Diamantas was expected to attend the announcement. Fedyukina told Axios that "there won't be any surprises for regulators in this."
The FDA's own guidance on using AI in drug decisions is still in draft. The agency issued draft recommendations in January 2025 on how sponsors should show that AI models used to support regulatory decisions are credible.
Fedyukina told Axios that there is no reason the program cannot target "a 10x reduction in the cost of clinical development for a drug," though she said it is hard to put a number on the possible savings. Those details, along with privacy protections, will shape whether patients trust the system enough to share their data.
Steps Patients Can Take Now
Patients do not need to do anything because of this announcement. ARPA-H is only now seeking proposals, so the program is not enrolling patients. People interested in current trials can search ClinicalTrials.gov or ask their doctor.
When any trial or app asks for health records, patients can ask who will see the data, whether it will be de-identified, how long it will be kept, and whether they can withdraw. Those questions apply whether a study is run by a university, a company, or a federal program, and patients can say no.
Key Questions Answered
What is SURPASS? It is an HHS and ARPA-H program that aims to use AI, modeling, and real-time data to make clinical trials faster and cheaper.
How much will it cost? HHS has not disclosed a budget. Reports say funded teams would receive five years of support.
Will my health data be used? Not yet. Companion projects plan national data connections and patient data sharing, but HHS has not detailed privacy protections.
Is the FDA involved? Axios reported that the FDA took part in planning. The HHS release does not describe a formal FDA role.
When are proposals due? Solution summaries are due Nov. 30, 2026, and full proposals are due Jan. 22, 2027.
What should patients ask before sharing data with a study? Ask who can access it, whether it is de-identified, how long it is stored, and whether you can withdraw.
Published by Medicaldaily.com
