AI is increasingly being used in clinical settings and has the potential to improve patient outcomes, but today’s payment models are limiting its adoption, according to a report released last week from the Peterson Health Technology Institute (PHTI).
To conduct the report, PHTI brought together leaders from health systems, health plans, technology developers, investment firms, academia and federal agencies in May 2026 to discuss payment options for clinical AI. Clinical AI supports the screening, diagnosis, treatment and management of health conditions, and includes assistive tools that help clinician decision-making and autonomous tools that perform clinical tasks.
Here are PHTI’s three key takeaways:
New Webinar: The Payer AI Readiness Gap: Why Better Data Will Define the Next Era of Health Plan Performance
We'll explore why the next era of payer performance will depend on building a trusted data foundation that connects people, providers, and relationships across the enterprise.
1. Applying today’s payment models to clinical AI will inflate costs: Today’s healthcare payment models mainly include fee-for-service (payment per visit, procedure or set of services), pay-for-performance (payment per service adjusted for quality or financial targets) and capitation (fixed payment per patient per period). However, none of these models were designed for clinical AI, and therefore, applying them to clinical AI will create a “structural misalignment,” the report stated.
“Under fee-for-service, reimbursement increases with the volume of billable services, rather than the value created,” the report said. “AI will enable healthcare organizations to deliver more services and generate more billable work at unprecedented scale. Layering this onto the existing fee-for-service payment chassis would allow reimbursement to grow far faster than the true cost of delivering that care.”
Performance-based and capitated models are better suited for clinical AI than fee-for-service, but they still lack the needed incentives to support adoption at scale, according to PHTI.
2. Payment models for AI should be deflationary, outcomes-based and evolve with evidence: PHTI’s workshop identified three principles for future AI payment models. One is that reimbursement should reward AI only when it improves outcomes, lowers costs or expands access.
Closing the Quality-Affordability Gap in Workplace Mental Health
In an interview, Kyan Health Co-Founder and Chief Commercial Officer Konstantin Struck discussed how Kyan gives mid-market and enterprise employers access to premium workforce mental healthcare, at a price point that is affordable.
In addition, payments should be tied to demonstrated clinical and financial value, versus just the volume of AI use. Lastly, the report states that payment models should evolve over time and adjust as evidence grows.
“Payment systems for AI should be more dynamic than those designed for traditional care delivery. Rates should start high enough to stimulate early innovation and adoption, then adjust as utilization scales, marginal costs decline, and real-world evidence accumulates—with regular, longitudinal review,” PHTI said.
3. Autonomous clinical AI needs entirely new payment models, not slight changes to existing models: The PHTI workshop participants reported that “no single payment model will support AI adoption across all clinical settings and use cases.” Therefore, new payment models for clinical AI need to be built for “specific contexts,” and need to address key questions like how prices should change over time, how coding and billing should evolve and who is eligible to receive payment for AI-enabled care.
“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,” PHTI said.
Photo: Issarawat Tattong, Getty Images