Dr. Anthony Chang

“We’re talking about the most powerful tools ever introduced into medicine, but they don’t yet fit into our economic architecture.”

Source unknown

AI in healthcare has shifted from experimental pilots to enterprise-level solutions with proven outcomes. Investment in AI spans three domains:

1) Operational Efficiency AI. Examples: Prior authorization automation, claims processing, predictive staffing. Economic impact: Reduces FTE needs, decreases turnaround times, lowers denials

2) Clinical Decision Support AI. Examples: Imaging triage, sepsis prediction, hospital-at-home triage. Economic impact: Reduces length of stay, prevents readmissions, improves capacity

3) Administrative and Revenue Cycle AI. Examples: NLP for documentation, AI-enabled coding, audit-prep bots. Economic impact: Increases coding accuracy, shortens revenue cycle, prevents revenue leakage

The cumulative economic impact is already significant. McKinsey estimates AI in healthcare could deliver $200–360 billion annually in savings in the U.S. alone.

Defining ROI for AI

ROI is defined as (Net Benefits – Cost of Investment) 

The traditional financial ROI may not be immediate for AI, but AI can deliver financial gains in specific areas in a relatively short timeframe:

ROI Dimension	Example	Financial Effect
1) Direct Savings	Automation of prior authorization	↓ Administrative costs
2) Revenue Uplift	NLP-enhanced documentation	↑ Billable services
3) Avoided Costs	Early detection of clinical deterioration	↓ ICU escalation, ↓ penalties
4) Productivity Gains	Automated scheduling or ambient scribe tech	↑ Provider throughput, ↓ burnout
5) Operational Efficiency	Scheduling, billing, and claims processing	↓ Administrative burden
6) Throughput Production	Triage and virtual assistants	↑ Capacity without adding staff
7) Strategic Positioning	Competitive advantage through AI leadership	↑ Market share, ↑ payer partnerships

 

For AI, therefore, ROI is more complicated than other technology services as the return can be non-linear and multi-faceted (especially with quality and safety dividends) as well as longer term. CFOs, therefore, must move beyond traditional capital expenditure or ROI metrics to include indirect and opportunity-based returns over 2–5 years.

Total Cost of Ownership (TCO) for AI

While AI promises efficiency and good ROI, costs are not limited to licensing. A clear-eyed TCO view includes:

1) Initial Costs: Licensing, infrastructure upgrades, integration with EHRs

2) Ongoing Costs: Maintenance, model retraining, monitoring

3) Change Management: Workforce training, workflow re-engineering

4) Compliance Costs: Privacy, security, FDA regulatory engagement. A successful AI investment must be grounded in cross-functional planning involving IT, operations, and clinical leadership—not just financial modeling.

Strategy for AI Adoption

The following is a basic strategic map for AI adoption in a healthcare organization.

1) Prioritize AI Use Cases with Clear Financial Outcomes. Start with low-hanging fruit: revenue cycle automation, documentation improvement, and predictive staffing

2) Demand Measurable KPIs from AI Vendors. Require vendors to articulate and contractually define ROI benchmarks: e.g., 15% reduction in denials, 20% documentation uplift

3) Stage AI Investments for Scalability. Begin with pilot projects, quantify returns, then scale enterprise-wide with a structured roadmap

4) Collaborate with Chief AI Officers or CIOs. Ensure financial oversight aligns with clinical and operational strategy. AI is not just a tech expense - it's a long-term asset.

AI is not just a cost center and it should be considered a revenue and margin expansion resource. The most successful healthcare executives will treat AI more like an asset class, not just a line item. ROI from AI is complex with non-linear and multi-faceted aspects and requires cross-functional leadership and time. Immediate gains are possible, but the biggest returns emerge with integrated AI portfolios deployed over 2–5 years. Just as AI has become multimodal, ROI on health AI should also be a convolution of financial, clinical, patient-centered, and social metrics.

The financial strategies for AI in healthcare will be among the popular topics covered at AIMed25, the world’s longest-running event dedicated to artificial intelligence in medicine and healthcare. Since 2013, AIMed has brought together over 1,000 clinicians across 50+ subspecialties, along with healthcare leaders, data scientists, trainees, investors, and innovators from around the globe.

This year’s 3-day meeting, taking place November 10–12, 2025 at the Manchester Grand Hyatt in San Diego, features dedicated tracks on AI in pediatrics and neonatology, AI in health professional education, and AI and mental health. Attendees can also expect hot-topic breakfast workshops, specialty breakout sessions, an abstract competition with scholar awards, the popular American Board of AI in Medicine (ABAIM) course, and—for the first time—a Chief AI Officer agenda.

Join us as we explore the future of healthcare, including the role of AGI, generative AI, large language models, intelligent XR, and much more. We look forward to seeing you there!