The Mount Sinai Health System is a New York-based healthcare network comprising multiple hospitals, outpatient facilities, and the Icahn School of Medicine at Mount Sinai. It provides a broad range of medical services across the city and conducts research and medical education alongside patient care. The system serves one of the most diverse patient populations in the United States.
You've spent much of your career focused on healthcare systems under pressure. What lessons from that work still influence how you lead Mount Sinai today?
My background is in emergency trauma, surgical critical care, and life-threatening systems like stroke and cardiac arrest. A central takeaway is that the underlying operating system of healthcare is as worthy of research as foundational science. For a long time, the industry focused almost exclusively on scientific discovery, which allows us to deliver extraordinary care today, but the next wave is transforming the delivery system itself.
That shift comes directly from watching health systems become stressed with access hurdles, navigation barriers, and poor patient experiences. Much of what you see in the tech industry right now is focused on rewriting this operational delivery system rather than lab science.
Healthcare data is notoriously fractured across different providers and health systems. How can modern technology solve this long-standing fragmentation and create a cohesive picture of a patient’s health?
We have made progress through point solutions—one-off apps for nutrition, exercise, or medications. However, these tools fragment care further because you end up with a dozen different platforms, none of which connect to your doctor. AI is powerful here because it excels at taking unstructured information, synthesizing it, and bringing it back into a unified, whole-person story.
The old vision for data sharing was top-down, where everything flowed into a single record system, but that has proven inadequate. The true unlock happens when the patient becomes the holder of their information, but giving patients total data control creates massive digital security vulnerabilities. When I was an ER doctor, identity fraud meant a patient using their cousin’s insurance card; today, that risk has migrated into sophisticated digital identity theft that health systems are unprepared to handle.
You were writing about telemedicine long before the pandemic. What did healthcare get right and wrong, and where do virtual care and AI fit into the future?
Trauma care taught me that moving decision-making virtually is different from moving physical procedures. If a patient needs an emergency abdominal operation, you need a helicopter to transport a surgical team. But when a patient needs immediate stroke or cardiac medication, you don't need to move the body; you just need the MRI images, CT scans, and physical exams to move virtually to a specialist who can coach the local team.
Today, virtual consultation is a commodity across subspecialties, moving seamlessly from hospitals into homes. The challenge now is that AI tools are democratizing medical information faster than payment systems, regulations, and accountability frameworks can adapt. We are moving into uncharted territory where commercial algorithms interact directly with patients, yet we still have no idea who is legally or clinically liable when a virtual agent serves up dangerous medical advice.
How should hospitals think about their role beyond the walls of the building, particularly around community health and social determinants?
This is an existential question about who ultimately owns a population's health outcomes—families, schools, churches, doctors, or tech companies. We have these conversations constantly because the largest dollar amounts are tied to healthcare, creating pressure to pull broader social responsibilities into the medical financial model. While we are deeply embedded in our neighborhoods, the solution cannot be to medicalize every social problem.
The boundary between traditional healthcare and community responsibility is blurred. Technology is changing this landscape, creating a generation of young consumers who are highly sophisticated about health and wellbeing, but we cannot ignore structural realities. Poverty remains the absolute worst of the chronic diseases, winding its way into physical and mental wellbeing, and solving it requires all of us rowing together.
How can advances in technology improve care access rather than widen disparities for diverse patient populations?
The future is already here—it is just not evenly distributed. In healthcare, if we build cash-pay longevity clinics or premium wellness services that are not covered insurance benefits, we risk widening existing disparities in the short term. However, if those innovations serve as testing grounds that eventually become covered, standard, and democratized for everyone, then we have created immense long-term value for the entire population.
A major policy reckoning is coming regarding how we manage populations that traditional health systems have left behind. Could low-cost, targeted agentic AI models eventually step in to act as primary care physicians for patients who currently have zero other options? Harm and mistakes will undoubtedly occur, but we always have to measure that risk against the counterfactual: if a vulnerable person has absolutely no access to care today, providing a digital alternative is a trade-off we must consider.
Where is Mount Sinai already seeing meaningful operational and clinical impact from AI?
We are structured to rapid-cycle learn across research, education, and clinical care, and we built the first Department of Artificial Intelligence and Human Health in a medical school to keep computer scientists focused on medicine. Clinically, the applications are staggering; in our NICU, digital surveillance models detect subclinical seizure activity in infants hours before it becomes visible to a human nurse. At the cutting edge, we are deploying brain-computer interfaces—laying a meshwork on the brain or using a smart stent—to allow paralyzed patients to type or move robotic limbs using only their thoughts.
Beyond these breakthroughs, AI is quietly revolutionizing business operations. We use predictive algorithms to alert rapid-response teams about patients at risk of deteriorating in their beds long before a crisis occurs, and we monitor tissue biomarkers to predict poor surgical wound healing weeks down the road. This is coupled with ambient listening tools to lower the administrative burden on doctors, alongside AI systems optimizing supply chains, operating room scheduling, and revenue cycles.
Mount Sinai recently partnered with Sophia Genetics. What problem are you trying to solve through genomics and precision medicine?
Two decades ago, the world was convinced that mapping the human genome would immediately explain everything about human health, but it didn't. We learned a great deal about how genes turn on and off, but genomic sequencing is still nowhere near a standard, foundational layer of medicine. True precision care cannot be achieved through lifestyle trackers alone; your smart watch can count your steps, but that is not precision medicine.
To get to the heart of individual health, we need deep, foundational building blocks. Your genetic sequence is a critical piece of that puzzle, but it must be layered with biomarkers that track chronic inflammation and how it constantly impacts brain health, cardiovascular disease, and overall longevity. Our partnership and the Mount Sinai Million initiative aim to make genotyping a routine part of standard care, unlocking true clinical personalization even as it forces us to confront difficult ethical questions about equity and access.
