Stanford Health Care is widely recognized not only as one of the top academic medical centers in the U.S. but also as a hub for cutting-edge innovation in clinical care, biomedical research, and digital transformation. The organization is uniquely positioned to redefine what technology-enabled healthcare can look like while navigating the high standards of ethics, governance, and patient trust that come with it.

In the second and final installment of this two-part interview series, Karl Hightower, Vice President and Chief Data and Analytics Officer at Stanford Health Care and the School of Medicine, speaks with Sezin Palmer, Health AI and Data Leader, Partner at EY, about the nuances of building an AI-powered healthcare ecosystem.

Hightower reflects on the ethical boundaries of emerging technologies, the promise of automation in decision-making, and the importance of creating an environment where both internal experts and external innovators can safely experiment and deliver next-generation impact.

Edited Experts

Q: As you explore different ways to use AI in healthcare, are there any areas, either based on Stanford policy or your perspective, where you believe AI shouldn’t be applied?

There are lots of things you do when it comes to understanding and knowing a patient. For example, when you walk into a retail store, facial recognition might be used to identify you and tailor the environment to your preferences. That’s where you need an ethics and governance board to help decide whether you even want to go down that path.

There’s a tremendous amount of technical possibility but the real question is, should you be doing it? That’s one of the things I appreciate about the leadership here at Stanford. They’re very thoughtful about how they want to apply these technologies, and the discussions around them are very open.

Historically, anytime you’re introducing new technology, AI, or any kind of data use, you need to ask: what does this mean for the organization? Should we be doing this? Those are hard decisions. They may not always align with the most immediate financial return, but you have to consider who you want to be in the long run.

Q: How do you view the current state of data and AI maturity in healthcare compared to other industries? What should the healthcare sector prioritize, and what do you think is achievable in the next three to five years?

Predicting three to five years out is going to get me in trouble. Just looking at the pace of GenAI, it’s hard to say with certainty where we’ll land. I could say that modeling and cleaning data might not even be necessary because those tasks could be automated tomorrow.

Even the way we do analytics today might change entirely, maybe we’ll just be talking to the computer, asking questions directly. There are so many possibilities. But when I think about it, I always come back to the basics: decisioning.

Why do we process all this data? Why do we build all these systems? It’s so that someone can make a decision and take action. That’s the core of it. So for me, it’s about simplifying, making sure we’re not overproducing or overengineering, but focusing on how people make decisions and what actions they take as a result.

Whether it’s building a complex model to evaluate images and text in a multimodal way, or just delivering a simple report with a key number, the focus has to stay on how that information is being used. And that’s where it gets exciting because in the next three to five years, not only will we have better ways to process information and arrive at insights faster, but ideally, using that information will become a whole lot simpler too.

Q: What steps are you currently taking or planning to take soon to bring your vision to life and make the most of these new capabilities?

No matter the industry, data is king when it comes to making this happen. I look at it as having all the data available at the right time to make whatever decision needs to be made. So, how much can I automate and take off people’s plates so they can focus on higher-level work? So they don’t have to spend as much time prepping things.

A lot of my early role is going to be about understanding how people use information, how they’re making decisions, and then getting as much automation in place as possible – so there’s a platform they can build from and deliver results quickly. It’s also about thinking ahead to what that next-generation technology is going to look like.

Q: How is Stanford approaching the development of these capabilities? Are you building them in-house, working with external vendors, or both? And if you are working with vendors, what do you expect them to bring to the table?

Stanford has a tremendous amount of internal talent, some of the smartest people in the world. And you go, “Okay, how do I enable them to do what they do best?” That’s where we look to partners – to take on some of the more foundational or routine tasks like moving data, cleaning data, and providing the platforms we need.

As we start exploring agentic AI, the orchestration of different agents and building the right toolboxes, we ask: is it one platform, is it multiple platforms, and how do we create a logical, safe environment for innovation?

It’s also about creating a space where external partners can bring in ideas and collaborate with our teams at Stanford to deliver the next generation of value. That kind of environment, especially given the amount of information we have, is far more powerful than just farming data and solving one-off problems. It’s about building compound interest, creating an ecosystem where innovation builds on itself.

Q: What single focus area should leading innovators prioritize to drive the greatest impact in U.S. healthcare?

If you look at innovation, you can break it down into two paths: am I just moving things incrementally forward, or am I trying to rethink the model, not just making the horse faster? It’s much harder to take a greenfield approach in healthcare, and that’s where you have to keep pushing through.

I hope regulations start to ease, just enough to make it easier to introduce new ideas. It’s also about giving people the freedom to try things. From a risk tolerance perspective, how do you offer those opportunities to patients? Creating that kind of environment would be amazing. But that will require some easing from the FDA and government regulators, so people can better understand and navigate the risks as we move into the next level of healthcare.

So then the question becomes: how do you help people innovate? It comes down to the environment. Healthcare is a tough space to break into – there’s a huge amount of data people don’t have access to, and a lot of nuance that’s not easy to grasp. That’s why we need spaces where people can come in, experiment, and start reshaping the healthcare system.

And I think that’s why you come to Silicon Valley. That’s why you’re here at Stanford because this is the type of environment that will change how healthcare is practiced.

CDO Magazine appreciates Karl Hightower for sharing his insights with our global community.