AI predicts and Medicare pays; a new model could change everyday healthcare

A quiet shift is underway in US healthcare policy, and it could reshape how artificial intelligence is used in medicine. A new Medicare payment model is creating room for AI-driven care in ways that were previously impossible, even as much of the broader tech industry remains unaware of the change.

At the centre of this development is Pair Team, a healthcare company focused on patients with chronic conditions who also face challenges such as unstable housing, food insecurity, or limited access to transport. Last month, the company was selected to participate in a new federal programme run by the Centers for Medicare & Medicaid Services, placing it among a small group of organisations testing how AI-enabled care can operate at national scale.

The programme, called ACCESS, marks a fundamental shift in how healthcare is paid for. Instead of reimbursing providers based on the number of visits or time spent with a clinician, the model rewards measurable health outcomes. Organisations receive predictable payments for managing conditions such as diabetes, high blood pressure, kidney disease, obesity, and mental health disorders, and earn the full amount only if patients show real improvement.


This matters because older payment systems only covered care during clinic visits, not what happens in between. There was no clear way to pay for AI tools that follow up with patients, manage referrals, or track medicines outside appointments. The new model changes that, making ongoing, tech-supported care financially possible.

For the Pair Team, this aligns closely with how it already operates. The company is based on the idea that health is shaped by more than medical care alone. It brings together medical, mental health, and social support, and relies on a large care team working directly with vulnerable patients. Studies show this approach can reduce avoidable hospital and emergency room visits.

Scaling this model, however, has historically required large human teams, making it expensive and slow to expand. Over the past year, the company has introduced a voice-based AI system to act as a constant point of contact for patients. Available around the clock, the system handles intake, follows up with patients, and helps coordinate care between visits. In many cases, it also provides something less clinical but equally important, which is, regular conversation and companionship for people who might otherwise be isolated.

The ACCESS programme itself reflects a more startup-style approach inside government. It was designed by former technology operators now working within the Medicare innovation arm, and includes features such as outcome-based payments, open competition among providers, and direct enrolment. The idea is to let different models compete, with better results determining which approaches succeed.

There are some hurdles, particularly around handling sensitive patient data and ensuring strong privacy safeguards within federal systems. Despite these challenges, interest is growing among healthcare investors, especially as funding returns to AI-led health startups. By changing how care is paid for, Medicare could be creating one of the clearest paths yet for AI to move from pilot projects to essential healthcare infrastructure.

Nominate now for ET AI Awards 2026