The National Institutes of Health is leaning into artificial intelligence to gather insights from a vast amount of health data — a shift that could allow the agency to conduct research more quickly and offer new tools to support clinicians.

Susan Gregurick, associate director for data science at NIH, said during Federal News Network’s AI & Data Exchange 2026 that advances in AI are beginning to unlock insights from data across disconnected systems.

“I think that there’s just an unlimited amount of excitement here,” Gregurick said. “I’ve seen some really exciting trends in developing AI technologies that can extract very difficult-to-find data from clinical records and notes that doctors take.”

AI to speed public health responses

At the core of NIH’s AI push is a push to break down data silos that have limited how quickly researchers can respond to emerging public health issues.

NIH is using AI in partnership with the Energy Department and the National Cancer Institute to extract information from pathology reports. Through this project, NIH is helping researchers get a better understanding of the relationship between COVID infections and cancer progression.

“The challenges we get really have to deal with getting real-time data about health,” she said. “It was very hard to understand the relationship between people who might have had COVID and the effect it would have on cancer progression,” Gregurick said.

“If you are a cancer patient, your diagnosis is going to come from a pathology report that’s in a completely different system than your electronic health record data. That program really does extract out the pertinent information from the pathology reports and then feeds it into our system for understanding cancer.”

Mining massive datasets

Beyond extracting insights from individual records, NIH is also grappling with the sheer scale of its data environment. NIH has about 440 petabytes of data across its three cloud service providers.

“This is really an impactful amount of data — getting our arms around all that data, discovering it, finding it, using it in analytics. That’s something that I’m very passionate about and spending a lot of time on,” Gregurick said. “There is an opportunity for using this sort of real-world data, like wearables and survey data, to help us understand health outcomes. Challenges persist in the structure of that data. Just getting it into standard formats from different devices is going to be an interesting amount of work.”

One example is the agency’s Bridge to AI program, which focuses on creating high-quality, AI-ready datasets.

“The whole goal of this program is to generate that AI-ready, high-quality, gold standard data that could then be amenable to new AI models,” she said.

NIH, through its Bridge to Artificial Intelligence (Bridge2AI) program, is generating new flagship datasets and best practices for machine learning analysis. These datasets were collected and processed with AI modeling in mind. One of the first datasets Bridge2AI focused on was the prevalence of Type 2 diabetes in American Indian and Alaska Native populations.

“We know that there’s a higher prevalence of Type 2 diabetes in those populations, and so creating the data helps researchers understand why these communities are at higher risk and how we can treat them,” Gregurick said. “This is one of the benefits of working together across NIH — creating that data and then making it available for researchers in populations such as American Indian, Alaska Native, or people who are from disadvantaged communities across the country.”

Using AI to improve back-office processes at NIH

AI is also playing a growing role in NIH’s internal operations, particularly in managing the approximately 20,000 grant applications the agency receives in a year.

“We’re using AI to do things like take grant applications that come in and cluster them into certain bins that would be appropriate for different study sections, so we don’t have to manually read the grants right off the bat,” Gregurick said. “We can use a large language model to cluster these into a relevant study section. And AI and large language models can also help us identify if are there any potential conflicts that reviewers may have with a grant, so they’re not assigned a conflicting grant, and help us select reviewers as well.”

NIH is also investing in the next generation of AI talent through programs like AIM-AHEAD, a community of over 10,000 AI and healthcare experts.

“The whole goal of AIM-AHEAD is to ensure that younger researchers, new clinicians or old ones and public health workers have access to and can get training on artificial intelligence,” Gregurick said.

Partnering with cloud providers

Partnerships with industry are another cornerstone of NIH’s AI strategy. Cloud providers have helped the agency scale its data infrastructure and provide training environments for researchers.

“The partnerships with AWS, Google and Microsoft Azure have been tremendous for us in terms of getting all of our data into cloud,” Gregurick said. “Each of the cloud service providers has built out a sandbox for students or for investigators to come and benchmark and try different algorithms, pull different datasets or just learn how to use the cloud to implement much higher throughput pipelines of analytics.”

AI is also enabling new types of real-time analytics and visualization, which she said could transform how health systems respond to emerging trends.

“If I wanted to know how many patients in the last 90 days have been diagnosed with lupus across our health network, having an AI-integrated dashboard with visualizations and analytics that actually can help us understand and pull that data in real time and show it to us is a game changer in terms of understanding larger trends,” Gregurick said.

Testing a health learning system in real time

The combination of AI, data integration and cross-sector collaboration is laying the foundation for a more data-driven health system.

“Working with startup companies or even larger ones, to really help us visualize these trends, is providing us insight that we just didn’t have before,” she said. “It’s allowing us to test what a health learning system that’s focused on patient information looks like in real time.”

While data challenges remain, NIH sees AI as essential to its mission moving forward.

“Gathering and harmonizing health data to identify trends, it continues to be a grand challenge,” Gregurick said. “We have a long way to go.”

But with AI increasingly embedded across research, operations and partnerships, she expects the agency can close that gap and deliver more precise insights into the nation’s most pressing health issues.

Discover more articles and videos now on the AI & Data Exchange event page.

Copyright
© 2026 Federal News Network. All rights reserved. This website is not intended for users located within the European Economic Area.