ARPA-H bets $62.7M on prescribing AI

Good morning, HealthTech insiders. The federal government just funded an ambitious clinical AI bet: developing autonomous AI agents for heart failure management that could assess symptoms, adjust medications, and generate prescription orders under clinical safeguards. ARPA-H committed up to $33.7 million in first-year contracts toward a $62.7 million program, requiring patient-facing teams to submit FDA authorization packages within 24 months.

At nearly the same time, federal regulators are changing both ends of the medical pipeline. Today, the FDA's reclassification of four oncology in situ hybridization therapeutic test systems from Class III to Class II takes effect, while the agency has launched its Expedited IND Pilot to shorten first-in-human trial preparation. Regulatory pathways are evolving, but the standards for clinical validation have never been higher.

In today's Signal:

• ARPA-H commits up to $62.7M to cardiovascular AI agents: Atman Health, Tempus AI, Updoc, Duke, Kaiser, and Stanford receive federal contracts to develop physician-supervised heart failure agents.

• FDA downclassifies four oncology companion diagnostics: Selected in situ hybridization test systems move from Class III PMA to Class II 510(k), effective September 16, 2026.

• FDA launches Expedited IND Pilot Program: HHS Operation TrialBlazer opens rolling pre-IND reviews aimed at strengthening domestic early-phase clinical development.

• Follow the Money: RegCell secures $66M for Nobel-linked Treg reprogramming while HexemBio announces $15.5M in total funding for stem cell rejuvenation.

In partnership with

Join 250,000 Founders & CEO's Staying Ahead

Ever reach the end of your workday and wonder what happened in the business world while you were busy running yours? Max’s CEO Report catches you up on the biggest business, tech, and AI stories in just a few minutes.

Each edition gives you the news and context worth knowing, so you can stay informed, spot new opportunities, and sound suspiciously well-read in your next meeting.

Sign Up Here For Free!

CLINICAL AI
1. ARPA-H commits up to $62.7M to develop prescribing AI agents for heart failure

The Advanced Research Projects Agency for Health (ARPA-H) announced contract awards under its ADVOCATE program, committing up to $33.7 million in first-year funding toward a four-year, $62.7 million initiative to develop autonomous AI systems for managing chronic heart failure patients.

The details:

• Autonomous clinical scope: Selected performers including Atman Health, Tempus AI, and Updoc are tasked with developing agentic AI designed to assess clinical symptoms, order lab tests, titrate guideline-directed medical therapy (GDMT), and generate prescription orders under clinical safeguards.

• Supervisory safeguards and clinical trials: Stanford University will construct a three-stage supervisory validation layer combining deterministic checks, predictive models, and AI review, while Duke University Health System and Kaiser Permanente lead prospective deployments across real-world health systems.

• Hard regulatory milestone: All patient-facing AI contractor teams must submit a completed FDA authorization package within 24 months of award, with the FDA collaborating on a regulatory framework for autonomous clinical agents.

The Signal: This contract represents a major federal effort to move from passive AI documentation assistants toward active, prescribing clinical agents. With nearly half of U.S. counties lacking a practicing cardiologist, ARPA-H is betting that algorithmic medication titration could prevent avoidable hospitalizations at population scale. By establishing a 24-month FDA submission target, alongside broader reimbursement discussions, the government is beginning to test the regulatory and payment infrastructure needed to integrate autonomous agents into clinical care.

REGULATORY INTELLIGENCE
2. FDA lowers barriers for four oncology companion diagnostics: ISH tests downclassified to Class II

The U.S. Food and Drug Administration's final administrative order officially takes effect today, reclassifying four in situ hybridization oncology therapeutic test systems from Class III Premarket Approval to Class II 510(k) premarket notification with special controls.

The details:

• Four key product codes affected: The order reclassifies post-amendment ISH and FISH companion diagnostics under product codes NYQ, MVD, OWE, and PNK, establishing a new codified regulation at 21 CFR 864.1890.

• Reducing review burdens and compliance costs: Shifting from a full PMA to a 510(k) pathway generally reduces regulatory burdens and avoids some PMA supplements, although significant assay modifications may still require new 510(k) submissions.

• Special controls replace PMA requirements: Manufacturers must adhere to standardized special controls, including analytical validation, cross-reactivity and cross-contamination testing, and multi-lot precision and reproducibility requirements. & Syndicate: The round elevates Inspiren's valuation above $500 million, with participation from Insight Partners, Primary Venture Partners, Scale Venture Partners, Vintage Investment Partners, and Lightbank.

The Signal: This is a meaningful structural change for precision oncology diagnostic developers. For years, the requirement to submit full Class III PMAs for these established FISH and ISH assays raised entry and update burdens and could delay companion test availability during targeted drug rollouts. By converting these four test types into Class II devices with special controls, the FDA is creating a pathway that could inform future modernization of other molecular diagnostics

CLINICAL DEVELOPMENT
3. FDA launches Expedited IND Pilot to strengthen domestic clinical development

The Food and Drug Administration officially launched the Expedited Investigational New Drug (IND) Pilot Program under HHS Operation TrialBlazer, opening applications through October 30, 2026, to help compress first-in-human trial preparation timelines in the United States.

The details:

• Rolling pre-IND collaborative reviews: The pilot establishes a rolling review structure where biotech sponsors partner with prospective Qualified Research Institutions, including academic medical centers, health systems, and specialized CROs, to resolve toxicology, pharmacology, and trial design issues before formal submission. Complete INDs remain subject to the standard 30-day review.

• Targeting 8 to 10 inaugural pairs: The FDA will select an initial cohort of 8 to 10 sponsor-institution pairs across a range of products and therapeutic areas, testing whether real-time regulatory interaction can shorten trial preparation timelines.

• Reclaiming domestic early-phase trials: The program responds to the trend of biopharma companies taking first-in-human trials to countries including China and Australia due to faster preparation and site activation.

The Signal: Early clinical development may be entering a faster phase. American biotechs have moved some Phase 1 trials abroad to reduce preparation and contracting delays. By testing rolling feedback with selected sponsor-institution pairs, the FDA is trying to make first-in-human trial preparation more collaborative while retaining the standard 30-day IND review.

👀 WHAT WE'RE WATCHING NEXT

1. ARPA-H agentic prescribing safeguards: We are watching the specific safety protocols Stanford and Duke establish to prevent flawed AI recommendations during diuretic and beta-blocker dose adjustments, particularly regarding lab turnaround latency and electrolyte monitoring thresholds.

2. Downclassification momentum for NGS multi-gene panels: We are watching whether the FDA's broader reclassification work eventually reaches nucleic acid companion diagnostics, including product codes OWD and PJG.

3. Sponsor-QRI selection dynamics: We are watching which sponsors and prospective Qualified Research Institutions apply as pairs before the October 30 deadline. The FDA plans to announce selected participants by December 18.

💰 FOLLOW THE MONEY

The deals:

• RegCell secures $66M Series A and grant extension for Nobel-linked Treg therapies: Epigenetic cell therapy biotech RegCell raised $66 million in total capital, combining a $44 million Series A with a $22 million grant from Japan's Agency for Medical Research and Development. Building on the Nobel-recognized discoveries of co-founder Professor Shimon Sakaguchi, RegCell reprograms pathogenic T cells into stable, antigen-specific regulatory T cells to restore immune tolerance.

• HexemBio announces $15.5M in total funding for stem cell rejuvenation platform: Longevity and synthetic biology startup HexemBio says its cumulative funding now totals $15.5 million, backed by investors including Draper Associates, Boost VC, and Happiness Capital. The company is developing a Synthetic Human Yolk Sac platform and machine learning to rejuvenate patients' aged hematopoietic stem cells for bone marrow transplantation.

• Veeva expands Falcon agentic AI for life sciences regulatory workflows: Veeva Systems continues the rollout of Falcon AI, integrating agentic medical, legal, and regulatory review automation across pharma compliance workflows.

The pattern: Federal capital allocation is visibly shifting toward autonomous clinical infrastructure and regulatory friction reduction. ARPA-H's ADVOCATE investment signals that clinical AI could move from passive administrative scribes toward high-acuity, closed-loop chronic care agents designed to improve guideline adherence. Concurrently, the FDA's Class II reclassification for four oncology companion diagnostics and the Expedited IND Pilot could reduce regulatory lead times for diagnostics manufacturers and early-phase biotech sponsors. Investors may increasingly focus on platform enablers providing deterministic guardrails, clinical validation infrastructure, and compliant workflow integration.

📆 UPCOMING HEALTHTECH EVENTS

• Rééduca Paris 2026 (September 17-19, 2026 | Paris, France): A leading European congress for rehabilitation, physical therapy, and assistive MedTech, bringing together approximately 6,000 professionals evaluating robotics and digital recovery platforms.

• World Medical Innovation Forum 2026 (September 22-23, 2026 | Boston, MA): Organized by Mass General Brigham and Bank of America, focusing on translating cutting-edge clinical AI, gene editing, and neurotechnology from academic labs into venture-backed clinical applications.

• Healthcare Excellence Through Technology (HETT) 2026 (September 29-30, 2026 | ExCeL London, UK): The UK's flagship digital health ecosystem conference, featuring NHS leadership, digital transformation directors, and healthtech founders navigating public procurement and AI safety frameworks.

🛠️ Trending HealthTech AI Tools

• Aletta Robotic Phlebotomy AI (by Vitestro): FDA De Novo authorized robotic blood-drawing device combining near-infrared imaging, Doppler ultrasound, and deep learning for autonomous venipuncture under trained phlebotomist supervision in outpatient settings.

• BioNeMo Generative Biology Platform (by NVIDIA): Enterprise-scale generative AI cloud foundation model suite enabling biopharma researchers to train, customize, and deploy molecular structure and protein interaction models at scale.

• QuantHealth Clinical Trial Simulation OS: Systems-biology AI simulation engine drawing on a company-reported dataset covering 350 million de-identified patient lives to predict drug-target interactions and optimize Phase 2/3 trial inclusion criteria before enrollment.

🛠️ Trending HealthTech AI Startups

• RegCell (Emeryville, CA | Led by Michael McCullar, CEO): Developing an antigen-specific regulatory T cell epigenetic reprogramming platform based on Nobel-recognized immunology for potential autoimmune and organ transplant therapies.

• HexemBio (Berkeley, CA and New York, NY | Led by Gabriel Levesque Tremblay, CEO): Synthetic biology and AI startup recreating embryonic yolk sac microenvironments with the aim of rejuvenating aged hematopoietic stem cells without genetic modification.

• Atman Health (Boston, MA | Led by Rahul Patel, CEO): Clinical AI developer selected for ARPA-H's ADVOCATE program to construct autonomous, physician-supervised agentic care models for cardiovascular disease management.

DECISION INTELLIGENCE

Highlights: Deep Dives & Industry Reports

• The Prescribing Agent Playbook: Inside ARPA-H's $62.7M Cardiovascular AI Mandate - A strategic teardown of how autonomous AI agents could enter clinical practice, the supervisory guardrails demanded by health systems, and the 24-month FDA submission roadmap.

• Breaking the PMA Bottleneck: How FDA's Class II Reclassification Affects Precision Oncology - An executive analysis of the 21 CFR 864.1890 final order, the unit economics of 510(k) companion diagnostics, and its potential implications for other molecular tests.

• Reclaiming Early Clinical Development: Inside the FDA's Expedited IND Pilot - How rolling pre-IND reviews and Qualified Research Institution partnerships under Operation TrialBlazer aim to shorten first-in-human trial preparation.

That's it for today!

Today's developments highlight a coordinated shift in how healthcare technology is developed, regulated, and delivered. The federal government is aiming to shorten clinical trial preparation through the Expedited IND Pilot while funding the development of prescribing cardiovascular AI agents.

At the same time, lowering regulatory barriers for four established diagnostics could help manufacturers bring companion tests to market faster.

The underlying questions for healthcare leaders are whether these technologies can perform clinical tasks safely at scale and how quickly health systems can adapt their governance, liability, and operating workflows to integrate autonomous capabilities responsibly.

See you tomorrow,

Dr. Dereck Mush, MD, MBA

The HealthTech Signal is independently funded. Sponsorships help us invest in better reporting, research, and a stronger newsletter for our readers. Want to reach HealthTech founders & executives while supporting the Signal? Get in touch.

Was this email forwarded to you? Sign up here.