Why providers still struggle to get paid for clinical AI

Current healthcare payment models fail to reimburse for clinical AI use

Most of today’s healthcare payment systems are stuck in the past. They were built around workflows where human clinicians provide care, log their time, and bill for each visit or service. Clinical AI changes that equation. It delivers continuous and scalable care, monitoring patients, making recommendations, and improving outcomes without direct clinician involvement every step of the way. The result is a structural mismatch: hospitals and providers invest in AI tools, but there’s no clear way to get paid for the value they create.

The Peterson Health Technology Institute (PHTI) report shows how this gap slows AI adoption, even when the technology improves quality and lowers long-term costs. Health system leaders are ready to expand AI use, especially in managing conditions like hypertension, but they’re constrained by payment structures that reward manual effort rather than digital efficiency. Without financial incentives aligned with AI-driven care, innovation gets trapped at the demonstration stage instead of scaling into everyday clinical use.

For executives, the takeaway is direct: current reimbursement models are not designed for continuous, autonomous systems. If this problem isn’t solved quickly, the healthcare sector risks missing a major opportunity to operate more efficiently and sustainably. Reform is business transformation at the infrastructure level.

Existing payment models inadvertently drive up costs or disincentivize AI adoption

The healthcare industry’s dominant payment structure, fee-for-service, has a built-in tendency to inflate costs when AI is introduced. By enabling clinicians to work faster and code more accurately, AI increases the number of billable events. According to research from America’s Health Insurance Plans and the Blue Cross Blue Shield Association, fee-for-service is still the most common method of payment. PHTI researchers note that revenue cycle AI solutions already drive a roughly 9% rise in medical costs by producing more billable codes. That’s inflation disguised as progress.

Value-based models, designed to reward better outcomes, also fall short. Pay-for-performance and capitation approaches don’t accommodate the profound workflow shifts required to integrate autonomous AI into care delivery. Accountable Care Organizations (ACOs), which manage total cost of care, face particular difficulty linking specific savings or quality improvements directly to AI tools. This weakens the business case for investment and adoption.

Executives should view this as a signal that the existing frameworks are too narrow for AI-driven care. The efficiency of AI challenges payment systems that rely on human labor as the main value driver. Moving forward, leaders must treat reimbursement strategy as part of their digital transformation roadmap. If models continue rewarding volume instead of verified outcomes, the industry will pay more while delivering less value, a direction that runs counter to the purpose of innovation.

A new, tailored payment model is essential for fostering clinical AI deployment

Healthcare needs a payment model that actually reflects the value AI brings. The current systems reward clinician time and procedural volume, factors that become less relevant when AI automates parts of care. Health system leaders contributing to the PHTI report emphasize that payment should depend on measurable improvement: better outcomes, lower total costs, or expanded patient access. Reimbursement should not hinge on hours logged but on the real impact of technology on health delivery.

A new model has to evolve with the technology. It should start by rewarding early adoption, providing meaningful financial incentives for organizations willing to take on AI’s upfront implementation costs. As clinical AI matures and its efficiency increases, reimbursement rates should gradually adjust to reflect lower marginal costs. Safeguards like price ratcheting, where payment rates decrease over time as operational efficiency improves, are critical to prevent inflated spending and ensure long-term sustainability.

For executives, the key is preparing for this shift now. As AI becomes increasingly autonomous, clinicians will move into oversight and quality assurance roles, focusing less on direct interaction and more on orchestration of AI-driven processes. That shift changes the economic logic of healthcare. Payment models must evolve to capture this new relationship between technology, labor, and outcome. Failure to do so will not only discourage adoption but also distort where and how clinical AI advances. Effective payment reform isn’t a policy exercise, it’s a structural adjustment to make high-value AI sustainable in real-world care.

Progress is underway through the development of new billing codes and pilot reimbursement models

Momentum is building. The American Medical Association (AMA) is creating new Current Procedural Terminology (CPT) codes specifically for autonomous clinical AI. These codes represent an important first step, they allow providers to bill separately for AI-driven services rather than including them under traditional physician codes. This aligns billing systems with the reality of AI: continuous, data-driven care that operates independently from hourly clinician time.

At the federal level, the Centers for Medicare & Medicaid Services (CMS) has taken a more experimental approach. Its Advancing Chronic Care with Effective, Scalable Solutions (ACCESS) Model is a 10-year voluntary payment program from the agency’s Innovation Center. The model replaces traditional fee-for-service billing with outcome-tied payments. Providers receive half of their monthly reimbursement upfront; the rest is issued only if specific clinical targets are met. This structure is designed to stabilize payments while prioritizing results over volume.

For executives, this development signals that the infrastructure for AI reimbursement is beginning to take shape. The AMA’s coding framework and CMS’s ACCESS pilot create a foundation for gradual adoption across the system. But adapting internal operations and documentation processes will be crucial. Organizations that understand these mechanisms early can position themselves ahead of the curve. The opportunity now lies in being ready for a future where healthcare payment centers on verified clinical value rather than outdated billing assumptions.

The ACCESS model, while a step forward, remains limited in scope and financial viability

The CMS ACCESS model marks solid progress toward integrating technology-enabled care, but its scope is too narrow and its reimbursement rates are too low to support meaningful adoption of clinical AI. The model focuses almost entirely on chronic disease management for Medicare beneficiaries. That limits its applicability to other specialties and excludes much of primary care, where many AI-driven efficiencies could emerge. For now, the range is too confined to drive large-scale transformation across the healthcare ecosystem.

Financially, the model struggles to meet the operational cost realities of AI-driven care. PHTI researchers and health system leaders point out that the ACCESS payment range—$180 to $420 per beneficiary annually, is not sustainable for providers offering remote monitoring, clinician oversight, and population health management. The rate is sufficient for digital health companies relying heavily on automation, but it underfunds the hybrid workflows that most clinical organizations still depend on.

Executives should view the ACCESS model as an early experiment rather than a final answer. It validates the idea that reimbursement can connect payments to outcomes and continuous digital care, but it remains unfit for broad use. Providers are hesitant because the payment gaps threaten operational stability, while digital health companies are optimistic because automation gives them lower cost structures. For corporate leaders planning AI integration, the lesson is to anticipate similar models that will evolve beyond chronic care and into general medicine. Early strategic alignment with these emerging frameworks will position organizations to capitalize when the payment environment catches up to the technology.

The future of AI reimbursement will affect both adoption speed and overall healthcare quality

The direction healthcare leaders take now in building AI reimbursement systems will determine how fast and how widely clinical AI becomes part of standard practice. Without a clear payment pathway, most organizations cannot justify the investment or the long-term operating costs. That delay slows innovation and limits patient access to more efficient care. PHTI researchers emphasize that addressing reimbursement isn’t just about fairness to providers, it’s about ensuring that AI can deliver on its promise to lower costs and raise care quality on a systemic level.

Healthcare is moving into a new phase where autonomous systems will underpin key operations, from diagnosis support to chronic condition management. The challenge isn’t whether AI can perform; it’s whether the ecosystem can pay for its use in a way that is rational and stable. Leaders who act early to engage regulators, test new payment models, and gather evidence on clinical outcomes will be better positioned as reimbursement structures evolve.

For executives, this is both a strategic and operational imperative. A comprehensive redesign, rather than incremental policy fixes, is needed to unlock the economic potential of clinical AI. Those who understand this shift early will not just adapt to change; they will shape it. The organizations that align financial models with technological progress will create the most sustainable and scalable foundations for the future of healthcare delivery.

Key takeaways for leaders

• Outdated payment models block AI adoption: Current systems pay for clinician time, not continuous AI-driven care. Leaders should push for reimbursement reforms that align payment with measurable outcomes and value creation.
• Existing structures inflate costs and limit innovation: Fee-for-service models drive unnecessary cost growth, while value-based ones lack incentives for AI use. Executives must reshape financial strategies toward outcome-based reimbursement to unlock AI’s efficiencies.
• A new model must reward results and sustainability: Future payment frameworks should tie reimbursement to proven improvements in care quality and access. Leaders should design systems that evolve with efficiencies, preventing overuse while maintaining long-term adoption incentives.
• Infrastructure changes are already in motion: The AMA’s new AI-specific billing codes and CMS’s ACCESS program mark early steps toward modernized reimbursement. Organizations should prepare by adapting internal billing, compliance, and outcome tracking systems.
• The ACCESS model is progress but needs expansion: Its chronic care focus and limited payment rates restrict scalability. Decision-makers should engage with policymakers to refine such models and advocate for fair reimbursement across broader care categories.
• Future reimbursement structures will define AI’s success: The pace of AI adoption depends on how quickly payment systems evolve. Leaders should act now to shape these frameworks, ensuring financial sustainability while advancing care innovation at scale.