Too often I hear healthcare leaders discuss health equity like an expensive add-on. We should be treating it as a financial asset.

A recent piece in JAMA Health Forum warns of the “inverse equity hypothesis” — the very real risk that AI in medicine will naturally benefit the most advantaged populations first. From an academic perspective, they are absolutely right to be concerned. But as an operational executive, I see the healthcare industry making a critical mistake in how we respond.

For those of us working to ensure every person and community has the opportunity to achieve their best health, we cannot wait for an invitation to the AI design table. Too often tech teams build and we audit. That passive approach is a mistake.

Guardrails only keep you from crashing. They don’t drive you to your destination.

We cannot settle for passive oversight; we must commit to active design. We must engineer AI to be an “Empathy Enabler” from day one.

I do not want AI to replace healthcare professionals; I want AI to handle the screens and the endless clicks so the multidisciplinary care team can finally look the patient in the eye and truly listen to them.

I do not view AI as a replacement for human connection. I view it as the ultimate tool to resolve the capacity tension in our clinics. Here is how we move this from a pilot program to an operational reality.

1. Erase the “Complexity Burden”

Expecting patients to just “figure it out” is costing the healthcare system at an alarming rate through unnecessary, avoidable spending. Forcing a patient to decipher complex medical jargon while they are sick or scared is the exact opposite of empathy.

By using AI to instantly adjust to patients’ preferences and specific health literacy needs, we can acknowledge and lean into what makes every person unique — their language, dialect, and culture — to automatically translate clinical jargon into plain, culturally resonant language.

2. Rescue the Care Team from “Fight-or-Flight”

In an era of acute workforce shortages, administrative weight inevitably spills over, leaving front desk staff, nurses, APPs, and physicians frustrated, burned out, and functioning in a constant state of fight-or-flight. Because of this, patients often become victims to a care team member feeling they simply do not have the time to explain or listen.

As the saying goes: Systems are perfectly built to get the results they get. We must build a new system. Let technology handle the administrative tasks. Let humans handle the connection.

3. Design With Patients, Not At Them

We can only realize AI’s full potential when people and patients are driving it. We cannot design systems assuming we know what is best — we don’t, because every person has a unique lived experience.

By bringing in diverse voices to provide cultural quality control, we prove to patients that their feedback matters. We must take a growth mindset with our technology, continually challenging it to do better and testing to ensure there are no hidden biases or exacerbations of inequity.

4. Trust is a Financial Asset

Academics often treat empathy as a “soft skill,” but in value-based care, trust is the currency of shared savings. When patients don’t trust the system or understand their care, they don’t — and often can’t — adhere.

  • A single avoidable readmission now costs an average of $15,200. Across Medicare, heart failure re-hospitalizations alone account for over $17 billion in annual spend.
  • A 2025 Nature Medicine study demonstrated that AI designed to bridge literacy and cultural gaps slashed readmissions by 18%, delivering a 3.4:1 ROI.
  • Hospitals that lead in patient experience and trust maintain net margins of 4.7% — nearly triple the 1.8% of their low-performing peers.

This is how we operationalize the Quintuple Aim — moving beyond cost-cutting to a model where clinician well-being, health equity, and financial sustainability are inextricably linked.

The Blueprint for the Future

We talk often in this industry about “meeting patients where they are,” but we need to design systems that require it. We shouldn’t just be “careful” with AI; we should be ambitious with it. We should challenge it to close the Social Drivers of Health gaps that humans have been too busy to bridge.

Technology handles the translation. Humans handle the connection. That is how we make high-quality, equitable care an operational reality for every community.

How are you quantifying ‘Trust’ in your 2026 value-based care models?