Nearly half of U.S. health systems are embracing generative AI, and the pace is only accelerating. For an industry often seen as trailing behind others in digital innovation, this momentum is a promising shift — driven by the growing realization that our data holds untapped potential to transform care.
As CEOs, we are all inspired by the promise of this groundbreaking tool. But the road to adoption is riddled with barriers, issues and mistrust, and there’s still widespread skepticism.
Despite all of this, many health systems are seeing promising results beyond the victories of enhanced clinical care and system efficiencies: about 60% of healthcare leaders who have implemented AI solutions are either seeing a positive return on investment or expect to. Operationalizing its use is a tremendous challenge. How do ideas become scalable pilots? How are safeguards guaranteed? How do we engage front-line teams?
Our AI journey is nearly three years old — here is what I have learned along the way:
- Be fully committed to taking a comprehensive approach. This requires detailed planning, breaking down silos, creating cross-functional teams and ensuring the highest ethical safeguards. Every health system should determine the best network-wide approach and select the areas most essential to its mission. You can’t boil the ocean; it’s important to prioritize areas of focus that match your mission and strategies.
At Hackensack Meridian, we created six areas of focus: disease detection, precision medicine, streamlining administrative and clinical processes, capacity management and burnout alleviation, creating personalized and equitable experiences, and enhancing research and innovation. We have more than 300 use cases in our pipeline. Many of them are in development, being piloted or deployed across our network. - Involve key stakeholders and create a culture and system where the entire enterprise is engaged. This means bringing physicians, nurses, other team members and trustees to the table. When we polled our board, their No. 1 priority for AI was enhanced disease detection.
Teams must understand the process to suggest new uses for AI. We created an effective process that starts with an email to the AI intake team, which includes domain subject experts who report directly to the chief AI officer. Often, the intake team will seek more information on the suggestion. Once more details emerge, the team has a very quick round-robin with technical experts. If they see a possibility, it will then go to an AI committee for review. If it passes muster here, it then goes to a domain expert, i.e. our chief medical officer and our working group. We take a deep dive into making the business case for the project. This whole process takes just three weeks on average. It’s a little like “Shark Tank” at this point — we ask very hard questions related to why we should invest in a concept. Again, we go back to the six areas of focus to determine potential value. - Build trust for internal and external audiences and demonstrate an unwavering commitment to ethics when using AI. This is vitally important. Healthcare professionals and consumers want to know where the information that AI-generated data has come from, who created it and how it was sourced. More than half of Americans say AI can help improve healthcare by reviewing important patient data, such as medical tests or diagnostic images. Among physicians, the use of AI for certain tasks nearly doubled in just a year — at 66% in 2024 — even as some doubts linger, according to an AMA survey released in February. These numbers will only grow.
AI technology is advancing so rapidly, can the regulations even keep up? It will be interesting to see the position the new administration takes on AI in the healthcare space. Vice President J.D. Vance recently noted the Trump administration believes AI will have countless revolutionary applications in economic innovation, job creation, national security, healthcare, free expression and more. What’s clear is that any governance should emanate from the healthcare sector itself, and we can work hand-in-hand with regulators in a collaborative partnership.
Here’s what’s coming next: we are headed into what is known as agentic AI. This is the future in which AI systems will start to tie together and perform multiple tasks. It can make decisions and take action. It can analyze situations and develop strategies. For example, an AI agent will take instructions from a patient and arrange appointments, transportation and more. This is just the start.
On this journey, we must always remember that humans are at the heart of healthcare. Technology cannot replace the human touch. And as we redefine and improve healthcare, we can never lose sight of our larger goal — enhancing the quality of people’s lives in the most fundamental ways.
Robert C. Garrett, FACHE, is CEO of Hackensack Meridian Health, New Jersey’s largest health network with 18 hospitals, more than 500 care locations, 7,000 physicians and the Hackensack Meridian School of Medicine.
