Medicare's ACCESS Model Opens First-Ever Government Payment Pathway for AI Patient Monitoring Agents

The Centers for Medicare and Medicaid Services (CMS) has finalized a payment model that creates, for the first time, a federal mechanism to reimburse AI agents performing continuous patient monitoring between clinical visits. The model, known as ACCESS (Advanced Community Care Coordination and Electronic Support Services), removes a fundamental barrier that has prevented health technology companies from deploying autonomous AI systems in routine care: there was simply no government pathway to bill for software that checks on patients, coordinates social services, or ensures medication adherence without direct clinician involvement.

The framework specifically targets what the industry calls "between-visit care"—the monitoring, outreach, and coordination work that historically required human staff but has increasingly become automatable. Under ACCESS, AI systems can be compensated for functions including automated wellness calls, housing and transportation referrals, and medication pickup verification. CMS designed the model with input from at least 15 health system partners, signaling a deliberate effort to align regulatory reimbursement with the operational realities of modern care delivery. The agency estimates the model will affect approximately 1.2 million beneficiaries initially across participating accountable care organizations.

The timing is significant: most major technology companies and health AI startups remain unaware of the model's existence and its implications, according to industry analysts familiar with federal health IT policy. Health systems that move quickly to certify AI tools under ACCESS could secure preferential positioning before capacity caps take effect. The model represents a potential inflection point for venture-backed health AI companies whose business models have relied on either direct-to-consumer pricing or pilot programs that never converted to sustainable reimbursement. CMS is expected to publish technical certification requirements by the end of the current quarter.