Medicare's New ACCESS Program Is Quietly Wiring AI Into Its Payment Architecture

By Hector Herrera | May 22, 2026

The Centers for Medicare and Medicaid Services selected 150 healthcare organizations—including AI-native startup Pair Team—for ACCESS, a 10-year federal program that tests outcome-based payment models explicitly designed around AI-driven care management. Unlike traditional Medicare fee-for-service, ACCESS providers receive flat monthly payments for managing patients with chronic conditions and only earn their full reimbursement when those patients hit measurable health outcomes. The program represents the federal government quietly restructuring how Medicare pays for care—and most of the health technology industry doesn't know it exists.

Why it matters: Medicare sets the financial rails that shape every other healthcare payment model in the United States. If ACCESS succeeds, it creates a direct economic incentive for AI-driven care coordination at scale—potentially transforming how health technology companies build products and how providers deploy them.

What Traditional Medicare Gets Wrong

Medicare's dominant payment model for most of its history has been fee-for-service: providers are paid a set rate for each procedure, test, or office visit. That model has well-documented problems. It incentivizes volume over outcomes. It rewards treating disease rather than preventing it. And it creates no economic signal for providers to invest in the kind of proactive, data-intensive care management that can reduce expensive downstream interventions.

Chronic disease management is where this failure is most costly. Patients with multiple chronic conditions—diabetes, heart failure, COPD, hypertension—drive a disproportionate share of Medicare spending. They are also the patients most likely to benefit from AI-driven monitoring that catches deterioration early, coordinates across care teams, and tracks social factors like housing and food access alongside clinical ones.

The Affordable Care Act created authority for CMS to run payment model experiments through its Center for Medicare and Medicaid Innovation (CMMI). According to TechCrunch, ACCESS is CMMI's most ambitious attempt to build payment incentives that reward exactly the kind of care that AI systems are suited to deliver.

How ACCESS Works

The model's structure is where the AI integration becomes visible:

• Target population: High-need, high-cost patients with multiple chronic conditions—the segment where AI monitoring and early intervention can produce measurable ROI
• Payment model: Providers receive a flat per-member-per-month capitation payment for care coordination, replacing the per-visit billing structure
• Performance incentives: Full payment is contingent on patients hitting defined health outcome benchmarks—reduced emergency department visits, controlled blood pressure readings, medication adherence targets
• Program duration: 10 years—long enough to generate real outcomes data, not just 18-month pilot results that healthcare innovators know are too short to mean anything

The 150 selected organizations represent a cross-section of care models: traditional primary care practices, community health centers, and AI-native companies like Pair Team, which uses AI to coordinate care for Medicaid patients across both clinical factors and social determinants of health—housing instability, food insecurity, transportation barriers.

Pair Team's selection is significant. It signals that CMS is not just accommodating AI-augmented traditional medicine; it is actively selecting organizations whose care model is built around AI as the coordination mechanism, not as an add-on tool.

Why Most of Tech Hasn't Noticed

TechCrunch's reporting frames ACCESS as a largely overlooked program despite its structural importance. Several factors explain the gap:

CMS moves slowly by design. Payment reform involves years of rulemaking, stakeholder comment periods, and regulatory review before money changes hands. By the time a model is selected and funded, the policy window that created it is years in the past—and media attention has moved on.

The AI angle is embedded, not labeled. ACCESS is not marketed as an "AI program." It is a payment model. The AI is in how selected organizations choose to coordinate care. Health policy reporters understand the payment mechanics; technology reporters understand AI. The story sits at the intersection that neither beat covers systematically.

150 participants is a small launch footprint. CMMI alternative payment models typically start small and scale based on outcomes data. The program currently affects a narrow slice of providers, reducing immediate visibility even though its implications—if it works and expands—are very large.

What This Means

For health AI companies: ACCESS creates a direct payment pathway for AI-driven care coordination. Companies that can demonstrate measurable patient outcome improvements within this model have a blueprint for scaling through Medicare—the largest single payer in the country, covering more than 67 million Americans. The program also gives AI health companies the 10-year runway needed to prove outcomes, rather than competing on 18-month pilot ROI.

For traditional health systems: Providers not selected for ACCESS should track its outcomes data carefully. If the model expands—as successful CMMI models often do—organizations without AI care management capabilities will face a structural reimbursement disadvantage compared to competitors who built those capabilities early.

For patients: AI-driven care coordination has shown promise in reducing unnecessary hospitalizations and improving chronic disease management in multiple research settings. ACCESS gives that evidence base a 10-year, 150-organization validation test in real Medicare populations. If it works, the result is better outcomes for millions of Americans managing serious chronic conditions.

For the broader AI industry: Healthcare is often cited as the sector where AI can deliver the most meaningful human benefit. ACCESS is the first federal program that puts a payment model behind that claim—not just research grants or pilot funding, but actual Medicare reimbursement tied to AI-coordinated health outcomes.

What to Watch

CMS will publish interim performance data on ACCESS participants beginning in year three of the program. Watch for those results—they will determine whether CMMI expands the model, restructures it, or terminates it. Watch also for whether incoming CMS leadership maintains CMMI's current innovation trajectory; value-based payment experiments have faced political headwinds before. And watch Pair Team and similar AI-native participants: their outcomes will be the most direct test of whether AI-first care coordination can outperform traditional models at Medicare scale.

Source: TechCrunch