Health systems are under intense pressure to scale artificial intelligence (AI). Yet in many cases, return on investment (ROI) and operational impact have not kept pace with spending or adoption. This is the paradox facing chief financial officers (CFO): AI investments are rising, but proven financial impact is elusive.
Last year, 27% of hospitals paid commercial licenses for AI applications compared to 9% of the overall economy. However, adoption by health systems is largely concentrated in incremental improvements — clinical documentation and other data-capture tools that bolster revenue cycle management (RCM) without fundamentally altering care delivery workflows or business models.
From an investment perspective, the start-up unicorn scene — companies with a valuation of over $1 billion — mirrors providers’ AI tool adoption, with 9 out of 11 of AI unicorn provider-facing companies concentrated in clinical documentation, revenue cycle management and diagnostic imaging, and decision support.
Despite the spending, health systems have yet to realize meaningful bottom-line impact from AI deployments, especially compared to other industries like financial services, media, and telecom.
As AI adoption expands, health systems must look beyond narrow deployments and ask: How could these tools generate tangible financial returns that impact operating margin?
Creating pathways to ROI
Financial benefit from AI tools can primarily be realized through four key paths: increase revenue per unit, reduce cost per unit, increase volume of care provided, and succeed in value-based care (VBC). While AI tools and vendors frequently claim to tap into different paths for value creation, the majority of the highly adopted solutions today realize value by increasing the unit revenue for care delivered by ensuring accurate and timely payments.
Importantly, implementing AI-driven tools alone will not improve the financial bottom line until implementation and downstream challenges such as clinical adoption and workforce restructuring are resolved. Deliberate operational, workforce, care model, or business model changes are needed to translate promises into financial reality.
There’s a consistent pattern across categories most impacted by AI today: the technology often works as advertised but does not deliver value when placed on top of fragmented operating models, processes, and policies. Organizations need to take an enterprise-wide view of their AI deployments. Below we outline areas where health systems can get value out of their investments.
Clinical documentation: Ambient scribing and AI-assisted coding can significantly improve note accuracy, reduce administrative burden, and better capture the intensity of services delivered. While this should enhance reimbursement integrity, as well as clinician satisfaction and productivity, gains are minimized if the operating model stays static. Health systems must drive sustained clinician adoption, redesign workflows so documentation efficiencies are actually realized, align coding and billing operations to capture improved specificity, and potentially rethink visit structure or team composition. If physicians continue seeing the same number of patients supported by the same staffing model, and documentation gains are simply absorbed as marginal time savings, the bottom-line impact remains limited.
Diagnostic imaging and clinical decision support: AI promises faster reads, improved accuracy, and earlier intervention. These capabilities can reduce downstream complications through such things as early sepsis detection or avoiding unnecessary testing. However, realizing financial benefit requires changes in clinical pathways and workforce roles. Radiologists and clinicians must incorporate AI outputs into decision-making. Care pathways must be redesigned to act on earlier signals. And organizations must determine how savings are captured: Will improved throughput allow higher imaging volume? Will avoided complications reduce costs in risk-based contracts? Without deliberate strategies to adjust staffing levels, redeploy capacity, or align incentives under value-based arrangements, improved diagnostic performance may not translate into measurable margin improvement.
Operational efficiency and workforce management: AI tools have the potential to optimize scheduling, staffing, and patient flow. They can identify bottlenecks, smooth capacity utilization, and improve throughput. Yet isolated optimization rarely produces sustained financial gains. To convert improved throughput into margin, systems must adjust labor models, close excess capacity, or strategically grow volume into newly available slots. AI tools in this category can help to more accurately forecast health needs at the individual, unit, and facility level, both long term and in real time.
Patient engagement and remote monitoring: Patient engagement and remote monitoring capabilities are especially powerful in managing patients with chronic diseases. But in fee-for-service environments, preventing admissions can reduce revenue rather than improve it. Financial upside often depends on participation in value-based care arrangements or other models that reward improved outcomes versus visits. Health systems must integrate remote monitoring into care management workflows, ensure provider engagement, and build the contracting structures that allow outcome improvements to translate into financial returns.
Three actions health systems should take now
First, systematically measure ROI beyond operational metrics. Tracking clicks saved, reduced documentation time, or improved denial rates is insufficient. CFOs and other stakeholders must connect AI deployment to unit economics: revenue per case, cost per encounter, labor per adjusted discharge, avoidable utilization under risk contracts, and ultimately operating margin.
Second, pair AI implementation with deliberate structural change. This includes workforce redesign, revised staffing models, care pathway standardization, throughput expansion strategies, and incentive realignment. AI should trigger operating model conversations not simply technology adoption.
Third, align AI strategy with business model evolution. Many AI use cases generate the greatest financial impact under value-based care, capitated arrangements, managed services, or new digital offerings. Health systems that remain structurally dependent on fee-for-service reimbursement may struggle to monetize tools that reduce utilization or improve prevention.
The CFO’s paradox is not that AI lacks value. It is that getting value out of the investments requires organizational commitment and discipline. Only when health systems combine AI capability with intentional operational and business-model transformation will today’s productivity gains translate into sustainable financial performance.