In May 2026, most of the American tech world is ignoring one of the more consequential AI stories of the year: Medicare’s innovation arm is standing up a payment model explicitly built for AI. ACCESS — Advancing Chronic Care with Effective, Scalable Solutions — is a 10-year program from the CMS Innovation Center. Its 150 selected organizations go live on July 5, 2026, and they will be paid for health outcomes rather than required activities. Pair Team, a chronic-care startup, learned it was accepted on April 30.
Why does this matter? Because reimbursement has always been the real bottleneck for AI in healthcare. However good the model, without a payment code it never enters the care pathway — and ACCESS builds that mechanism for the first time.
What ACCESS Is
ACCESS is structured as per-member payments with outcome bonuses: organizations receive a predictable monthly fee per member, but only collect in full when patients hit measurable goals — lower blood pressure, reduced pain. Coverage targets chronic conditions: diabetes, hypertension, chronic kidney disease, obesity, depression, and anxiety.
The program’s architects are unusual for CMS: Innovation Center director Abe Sutton, formerly a venture investor, and Jacob Shiff, chief AI and technology officer and a former healthcare founder. The first cohort of 150 includes AI-doctor startups, virtual nutrition providers, connected-device firms, and wearable makers such as Whoop.
Why the Old Payment Model Couldn’t Fit AI
Traditional Medicare reimburses clinician time. An AI agent that monitors patients between visits, coordinates housing referrals, or confirms a prescription was picked up has no line item to bill against. As TechCrunch put it, ACCESS “creates that mechanism for the first time.”
Pair Team CEO Neil Batlivala calls it “a payment model transformation”: “You just couldn’t do this before.” It also aligns with where the clinical evidence has been heading — value that shows up in how patients interact with systems over time, much like the patterns documented in emergency-room AI diagnosis studies (/blog/harvard-ai-er-diagnosis-study/).
Pair Team and Flora: The Case Already Running
Pair Team, founded in 2019, serves chronic-condition patients facing housing instability, food insecurity, or transportation gaps — roughly a third of Americans. The company fields about 850 clinical professionals, claims California’s largest community health workforce, reports revenue above nine figures, and has raised around $30 million from Kleiner Perkins, Kraft Ventures, and Next Ventures. A peer-reviewed study in the Journal of General Internal Medicine found its care reduced avoidable ER and inpatient use; by Batlivala’s accounting, one in four hospital visits and one in two ER visits simply never happen under its care. Its partnerships reach roughly 500,000 potential patients, with a stated goal of one million within three years.
About nine months ago the company deployed Flora, a voice AI agent handling intake, referrals, and check-ins around the clock. The pivotal case: a 67-year-old woman living in her car with PTSD and congestive heart failure who spoke with Flora for over an hour. Batlivala called it “both incredible and depressing” — “That’s the companionship piece,” and “it turns out that is truly an intervention.”
Low Rates Are the Design, Not a Flaw
Per-patient monthly payments came in lower than many participants expected, and Batlivala argues that is intentional: “the reimbursement rates have to be low,” because “the economics only work if you’re running a lean, AI-first operation.” In other words, ACCESS is not subsidizing labor-intensive care delivery — it is forcing participants to put AI at the core of their cost structure.
The Risks: Privacy and the Innovation Center’s Track Record
Two risks deserve scrutiny. Privacy: sensitive patient data will flow into federal infrastructure with a breach history — CMS previously leaked providers’ Social Security numbers through a directory. Money: a 2023 CBO analysis found the CMS Innovation Center increased federal spending by $5.4 billion over its first decade. “Savings” is a hypothesis to be tested, not a promise.
The backdrop is capital returning to the sector: Q1 2026 was digital health’s strongest funding quarter since the pandemic, driven mostly by AI. Once the payment side is unblocked, this cohort will move faster than oversight can keep up — and the 10-year ACCESS clock is the timer on that experiment.
Sources
- Medicare’s new payment model is built for AI — TechCrunch
- ACCESS Model — CMS Innovation Center
- Pair Team accepted into CMS ACCESS model — PR Newswire
AI-assisted summary compiled from the sources above, reviewed by a human before publishing.