The Intelligence Economy: Breaking Health Care’s “Iron Triangle”

Key takeaways

• Health care is an overlooked AI beneficiary: We think the market is giving little credit for the benefits of AI in health care.
• Breaking the Iron Triangle: Scalable medical intelligence can improve quality, access and cost at the same time, rather than trading one for another.
• AI is a force multiplier for health care innovation: AI accelerates advances already underway across drugs, devices, diagnostics and care delivery.

Health care is one of the largest sectors in most developed economies—yet it remains one of the least recognized beneficiaries of artificial intelligence (AI). We think the market may be underestimating AI’s potential to reshape the sector. To understand why, it helps to revisit the "Iron Triangle", a decades-old framework for health care systems that focuses on the constraints between quality, access and cost.

Health Care Has Rarely Mattered More to the Economy—or Less to the Market

Early evidence is emerging that both the health care sector and the patients it serves will benefit, perhaps substantially, from advances in AI. Yet, we believe today’s investors are assigning little value to that long-term potential.

By the end of 2026, health care is expected to account for over 18% of the US economy, yet it currently represents only 9.1% of the S&P 500.¹ That gap highlights a striking disconnect between the sector’s economic importance and its market value, just as AI appears poised to begin reshaping its long-term growth potential.

We see attractive investment opportunities among health care companies where ongoing innovation trends, increasingly amplified by AI, can strengthen long-term earnings power. Because most health care companies aren’t priced as near-term AI beneficiaries, a select few may offer both diversification from AI-driven volatility and potential upside as AI improves their underlying business models over time.

But the potential benefits of AI in health care extend beyond individual companies. Early proof points suggest AI can fundamentally rebalance the Iron Triangle's long-standing trade-offs.

A Decades-old Constraint: Health Care’s Iron Triangle

In Medicine’s Dilemmas, published in 1994, William Kissick described health care as a give-and-take process between three critical variables: quality, access and cost. Kissick used a triangle to symbolize these trade-offs with each side representing one variable and each variable constraining the others. The idea reflects what we have all experienced:

• Broader access raises spending and / or reduces quality.
• Increased quality involves accessing a finite number of specialists with above-average fees and a penchant for expensive technology and medicines.
• Aggressive attempts to control costs invariably restrict access to these same higher-quality specialists, drugs and technologies.

The triangle is “iron” because Kissick saw few ways to change the restrictive compromises between quality, access, and cost.

The trade-offs look different by system and country but are present everywhere: single-payer systems tend to maintain cost and quality by rationing access through waiting lists; the US maintains access and quality for the insured at exceptional cost; and many emerging markets lack the specialist supply to deliver quality at any price.

AI May Improve Access and Cost Before It Revolutionizes Drug Development

Innovations in drugs, devices and diagnostics have traditionally improved the quality side of the triangle through greater efficacy and more personalized care. Recent advances in AI could accelerate breakthroughs already underway across cancer, cardiovascular disease, autoimmune disorders and other areas. But meaningfully improving success rates in drug development may still be years away given the extraordinary complexity of human biology, limited high-quality training data, and the need to validate results in years-long human trials. In the interim, the more transformational opportunity for AI may lie on the other two sides of the Iron Triangle: access and cost.

The triangle was iron for a reason: medical intelligence was scarce. It lived inside a finite number of experts, buildings, and 15-minute appointment slots, so expanding access to it meant spending more or diluting quality. AI changes the underlying constraint. When specialist-grade intelligence can be delivered through a coach, a nurse, or an app at low marginal cost, expanding access no longer requires expanding spending—and continuous, low-cost access unlocks the cheapest form of care there is: prevention.

The opportunity to improve access is significant given more than 100 million Americans live in communities without adequate primary care.² Demand far exceeds supply—many physicians are retiring, burnout remains high, and fewer than half of graduating medical doctors express an interest in primary care.³

AI-infused hybrid care models are beginning to close this gap

By combining AI with nurse practitioners, health coaches, and smartphone apps, these models put specialist-grade medical intelligence in patients’ hands around the clock—in ways that build patient trust. Many patients now prefer these low-cost, continuous relationships to waiting months for a primary care physician or specialist appointment.

Better access, in turn, lowers cost. When patients can receive quality medical guidance early and often, chronic conditions like diabetes and hypertension can be managed before they spiral into complications that cost tens or hundreds of thousands of dollars to fix. 

AI can also help address another source of scarcity: clinician time. Doctors and nurses spend an estimated 20%–30% of their time on administrative tasks.⁴ Automating even a portion of that work could return scarce clinical capacity to patient care and help ease labor constraints.

These benefits reinforce one another. Savings can be banked—or reinvested in greater access to higher-quality drugs, devices and diagnostics, where AI can further amplify an already healthy cadence of innovation.

How AI-infused care can improve quality, access and cost for a Medicare patient with chronic conditions

The Stakes Are High in Health Care, but Systemwide Trust Is Low

Most health systems, and the US fee-for-service model in particular, often pay to fix us after we break, not to keep us from breaking in the first place. Medical intelligence has historically been locked inside appointment slots—at a specialist’s office, a hospital, or a laboratory—and reaching it means navigating an expensive, disjointed maze of health care providers. 

Some companies are now pursuing a different approach with AI-infused hybrid care that bypasses dysfunctional legacy infrastructure and separates intelligence from place of service. The result improves quality, increases access and lowers cost—simultaneously. 

These companies are using AI systems to build trust with patients. The adage that “no one cares how much you know, until they know how much you care” carries extra weight in health care, where the stakes are high, but systemwide trust is low. Patients are unlikely to embrace new innovations if they don’t trust the providers recommending them—and trust is hard to build when patients see their doctor for just 15 minutes a few times a year.⁵ When AI absorbs the paperwork and helps guide interactions, coaches and nurses gain the time to know a patient’s life and connect the care plan to what that patient truly wants—whether that is staying independent or dancing at a granddaughter’s wedding. And when trust runs high, patients often grant providers more freedom to innovate.

Trust fell by almost half in under four years

Share in US adults expressing "a lot" of trust in physicians and hospitals

AI Is a Force Multiplier for Innovation Through Recombination

Most innovations involve recombining existing ideas with new ones to arrive at novel solutions. To illustrate the power of recombination, consider the four established approaches to treating cancer: surgery, radiation, chemotherapy and immunotherapy. The combination of these approaches results in 24 total treatment options for a given cancer (4x3x2x1). 

In August, Moderna and Merck announced that the Phase 3 trial of their first-in-class mRNA cancer vaccine, known as V940, succeeded in the treatment of skin cancer (melanoma).⁷ Used in combination with Merck’s immunotherapy, Keytruda, V940 demonstrated statistically significant and clinically meaningful improvements in recurrence-free survival and distant metastasis-free survival compared with Keytruda alone.  

If the cancer vaccine developed by Moderna and Merck ends up representing a new approach for treating cancer, then society’s treatment options didn’t simply increase by one, they increased by a factor of five, to 120 (5x4x3x2x1).⁸ Adding a new treatment modality expands the number of ways existing approaches can potentially be combined to create exponentially more options. This is the power of recombination as a central driver of innovation across not only therapeutics, but also specialty diagnostics, medical devices, and countless industries outside of the healthcare sector. 

AI adds another dimension: It can accelerate the discovery of new ideas and efficiently identify the best possible solutions that emerge from an ever-expanding set of possibilities.

The Market Is Giving Little Credit for the Benefits of AI in Health Care

We see clear early signals that AI-infused health care solutions can break the historical constraints between health care quality, access and cost.

Regulations and misaligned incentives may slow the pace of adoption, but the direction of travel is clear and the magnitude of impact looks significant. 

All investments can be thought of as some combination of price paid (what you give) and future earnings power (what you get). At current prices, we believe investors are giving little credit for

AI-driven innovations that can increase the long-term earnings potential of many health care companies in the US and around the world.