Summary
- Applying today’s payment options to clinical AI will both inflate costs under fee-for-service and insufficiently incentivize adoption under pay-for-performance and capitation models.
- Payment models for AI should be deflationary, outcome-based, and adaptive to evidence and an evolving clinician role.
- Autonomous clinical AI requires new payment models, not incremental modifications to existing ones.
Key Takeaways
In May 2026, PHTI convened senior leaders from health systems, health plans, technology developers, investment firms, academia, and federal agencies to discuss payment that supports the adoption of high-value clinical AI. The workshop centered on current and future payment models for assistive and autonomous AI, focusing on hypertension management to illustrate how care delivery and payment could evolve as AI takes on a greater role in clinical care delivery.
Participants gathered to explore the following questions:
- Which existing payment options can be adapted for increasingly autonomous AI delivery models?
- Where do payment options fall short and where are incentives misaligned?
- How should future payment options be designed to support high-value clinical AI adoption?
Applying today’s payment options to clinical AI will both inflate costs under fee-for-service and insufficiently incentivize adoption.
Today’s healthcare payment models—fee-for-service, pay-for-performance, and capitation—were not designed for technologies that can independently perform clinical work. As a result, applying today’s payment structures to clinical AI creates a structural misalignment.
- Under fee-for-service, reimbursement increases with the volume of billable services, rather than the value created, which would enable healthcare organizations to deliver more services and generate more billable work at unprecedented scale.
- Performance-based and capitated models are better aligned than fee-for-service payment but lack design features required to drive adoption at scale.
Payment models for AI should be deflationary, outcome-based, and adaptive to evidence and an evolving clinician role.
Workshop participants identified three core principles to guide the design of future payment models for AI-enabled care:
Autonomous clinical AI requires new payment models, not incremental modifications to existing ones.
New payment models for clinical AI must be deliberately designed for specific contexts—no single payment model will support AI adoption across all clinical settings and use cases. Key questions will need to be addressed across several core dimensions, such as how prices should be set and change over time.
Clinical AI presents a once-in-a-generation opportunity to fundamentally redesign how healthcare is delivered. Unlike previous new technologies, clinical AI has the potential to reshape the underlying economics of care itself by increasingly substituting technology for human labor. If deployed thoughtfully and at scale, this could dramatically expand access to care, improve patient outcomes, and lower healthcare costs over the long term.
The decisions we make today on how to pay for AI-enabled clinical care will shape not only the pace of clinical AI adoption, but the impact on our healthcare system for years to come.
AI Applications in Healthcare
PHTI convened a three-part workshop series in Washington, D.C., bringing together senior leaders from health systems, health plans, technology developers, academia, investment firms, and federal agencies to explore pathways for responsible AI adoption across healthcare.