Medicare’s new payment model is built for AI, and most of the tech world has no idea

DeepTrendLab's Take on Medicare’s new payment model is built for AI, and most...

The Centers for Medicare & Medicaid Services has just quietly opened a door that regulatory bodies rarely unlock for experimental technology. In accepting 150 companies into ACCESS—Advancing Chronic Care with Effective, Scalable Solutions—the federal government is running a decade-long pilot program that fundamentally reframes how AI can be integrated into one of America's most hidebound institutions. The program launches July 5, and its participant list includes everything from AI-native doctor startups to wearable makers, but the real story isn't the companies selected. It's the payment mechanism that makes them all possible. For the first time, Medicare is creating reimbursement for AI systems that don't fit the traditional billing model—the agent that monitors between visits, the system that coordinates social services, the algorithm that prevents readmission. This is structural change masquerading as a pilot.

The healthcare system has been locked in time-based payment for decades. A clinician's reimbursement depends on contact hours, not outcomes, which created a misalignment between what medicine can actually do and what the payment system incentivizes. AI disrupts that math entirely. An AI agent handling chronic disease management generates value that traditional billing codes can't capture. It doesn't require a clinical encounter; it can work asynchronously, scale across populations, and coordinate care across fragmented systems. Yet the regulatory and financial infrastructure had no way to pay for these capabilities. Pair Team, the deep player in this space, spent five years building infrastructure for patients with chronic conditions complicated by housing instability, food insecurity, and transportation barriers—exactly the population where coordinated, asynchronous care creates the most value. ACCESS exists because policymakers finally recognized that the regulatory bottleneck was the payment model itself, not the technology.

What's happening here extends far beyond healthcare. By creating "swim lanes" for AI innovation inside a traditionally regulated industry, CMS is signaling that innovation in healthcare won't require overturning existing regulation—it requires rewriting the incentive structures that regulation enabled. This is a direct response to a broader Silicon Valley frustration: the assumption that compliance and creativity are zero-sum games. In healthcare, they don't have to be. The government is essentially saying: build what works, prove the outcomes, and we'll find a way to pay for it. That's radically different from the traditional regulatory approach, which defines what's allowed and forces innovation into those boundaries. For a platform like CMS, which controls hundreds of billions in annual spending, this pivot is seismic. It means that the next decade of health-tech innovation won't happen around the regulatory structure—it will happen because the regulatory structure was reshaped to accommodate it.

The immediate beneficiaries are health-tech developers who've been watching the disconnect between what's possible and what's reimbursable. For early-stage founders, ACCESS suddenly makes business models that were previously impossible viable at federal scale. But the program also forces a reckoning on what "innovation" means in healthcare. The cohort includes connected device companies and wearable makers whose value proposition appeals to wellness-conscious consumers—the exact population that isn't struggling with social determinants of health. Batlivala's pointed skepticism about wearables for seniors dealing with food insecurity isn't cynicism; it's precision. The companies that will actually move the outcomes needle in ACCESS are the ones solving coordination problems and addressing root causes, not optimizing for the worried well. This distinction matters enormously for the trajectory of health-tech investment over the next decade.

The competitive dynamic here reveals a widening chasm in how different sectors of healthtech think about the problem. Consumer-facing companies and AI researchers focused on clinical accuracy are optimizing for entirely different variables than companies built around social determinants. Pair Team's nine-figure revenue and 850-person workforce suggest that the market is already sorting toward the latter—but most of the VC and media attention still flows toward the former. ACCESS could accelerate that realignment, but only if the program's outcome data proves decisive. That's the real question: when CMS publishes the results, will the companies that addressed the full context of patients' lives actually outperform the ones that optimized for clinical metrics alone? If they do, the investment thesis for health-tech will shift materially. If they don't, we'll have learned that government-scaled AI in healthcare works differently than anyone expected.

Watch how quickly other regulated industries attempt to copy this model. Insurance, financial services, and education all have similar misalignments between what technology can do and what existing payment structures allow. If ACCESS generates clear evidence that outcome-based payment unlocks innovation without compromising quality or safety, this becomes a template. Also watch the failure rate. Not all 150 participants will succeed, and the ones that fail will reveal whether the bottleneck was genuinely the payment model or whether some players simply lacked the operational depth to execute at scale. Most importantly, track whether ACCESS outcomes actually correlate with companies that addressed social determinants versus those that didn't. That data will determine whether the government learned to write checks for solutions that work, or simply found a new way to subsidize incrementalism.