928: Approachable AI for Public Health

SUMMER JOHNSON: 

This is the award-winning Public Health Review Morning Edition for Monday, June 16, 2025. I'm Summer Johnson. Now, today's news from the Association of State and Territorial Health Officials.

TRUC TAYLOR: 

At the end of the day, AI is simply just another tool in the public health toolbox, but it's one with enormous potential if we approach it thoughtfully. We don't have to wait for perfect systems. We can start small, learn, and build the organizational capacity now.

JOHNSON: 

Public health leaders are interested in using AI effectively and ethically. So today, we have a special episode to help departments approach AI with confidence. Truc Taylor is the director of health data analytics at Guidehouse, a founding member of ASTHO's Innovation Advisory Council, a new multisector collaborative of private sector partners, keeping ASTHO's membership and leadership informed and nimble on emerging topics affecting public health.

Truc, we have a lot to cover today on what a lot of people assume is a very complicated topic, so let's get right into it. What do you believe is really the most compelling reason to use AI as public health leaders?

TAYLOR: 

I often tell public health leaders to think of AI not as a mysterious black box, but as a new colleague. So, one who can shift through massive amounts of data in seconds, spot patterns we may overlook, and suggest options that help us make better decisions. In my opinion, the hook is simple: AI can give us better, faster situational awareness, whether that's in tracking outbreaks, predicting resource needs, or identifying vulnerable populations. It can even help us do our jobs faster, like grant writing or automating workflows. When you realize that AI isn't about replacing expertise, but about amplifying it, the technology can become less intimidating.

JOHNSON: 

How important is it to keep a human in the loop, as we've heard a few times before?

TAYLOR: 

In public health, the human in the loop isn't optional. It's essential. AI can process complex data and suggest insights, but public health decisions have real human and population consequences. Context, ethics, equity considerations, and cultural factors require human judgment that AI cannot replace and replicate. And critically, many of the data sets we rely on in public health are messy, incomplete, or contain historical biases, whether that's in case reporting social determinants or access to care without a human oversight, AI models risk perpetuating or even amplifying those biases. Think of AI as a decision support tool that can help surface correlations, flag anomalies and predict likely scenarios, but it's the public health professional who must assess whether the patterns make sense, whether additional data is needed, and whether there may be any unseen confounding factors. The most effective implementations of AI and public health are ones where humans and machines collaborate and where humans remain ultimately accountable for interpretation and action.

JOHNSON: 

How can organizations build competence in AI systems, and what can we do to ensure that the answers provided by AI are reliable, and accurate, and trustworthy?

TAYLOR: 

Trust is earned through transparency, validation, and ongoing oversight. Period. Public health organizations should prioritize explainable AI models that allow practitioners to understand why a recommendation is being made, not just what the outcome is, but even explainable models need human validation. Right data sets may contain biases, such as underrepresentation of certain communities or historical inequities and access to care, and if those biases aren't recognized and controlled, AI can inadvertently reinforce them. That's why multidisciplinary oversight is so important, bringing together data scientists, epidemiologists, ethicists, and even community representatives to catch blind spots and course correct when models make mistakes, which they inevitably will. AI systems also need ongoing recalibration as public health conditions and data sources evolve. What works today may not work next year, especially as novel public health events emerge. Right? And this isn't just theory. So, according to a recent study that Guidehouse and CDO Magazine conducted of public and commercial technology leaders. In that survey, 76% of organizations support that they are not fully equipped to govern and oversee AI effectively. Now, that gap highlights why agencies need to build governance structures first.

JOHNSON: 

Do you have any examples of recent public health work that has successfully incorporated AI into its processes? And what made those examples successful?

TAYLOR: 

Absolutely. So, one strong example is how several health departments have used AI-driven syndromic surveillance during COVID-19 to detect early signals of outbreaks from emergency department visits, even using social media and wastewater surveillance data as well. So, these systems helped identify local surges earlier than traditional reporting did, and allowed faster public health insights and reactions. Now, even CDC's Legionnaires' Disease Team is using AI to automatically detect cooling towers from aerial imagery to accelerate their ability to respond to outbreaks. Pretty cool, right? Another example is the use of AI-powered natural language processing, or NLP for short, to extract insights from unstructured clinical notes or social service records to identify social determinants of health.

JOHNSON: 

If you're speaking to someone or an organization that wants to incorporate AI into their public health processes, what are some steps that you tell them they can take?

TAYLOR: 

Great question. I start by saying, begin with the problem definition, not the technology. Look for areas where you have complex, data-rich challenges that really strain from your current resources, whether that's improving case investigation workflows, optimizing resource allocation, or identifying populations at higher risk. So, one great example is what Kansas' local health departments are doing. They've developed an AI integration roadmap that lays out a deliberate process for adopting AI responsibly, building internal capacity, addressing workforce development, ensuring strong governance, and keeping equity and community trust at the center.

JOHNSON: 

What would you like to leave your colleagues with today?

TAYLOR: 

At the end of the day, AI is simply just another tool in the public health toolbox, but it's one with enormous potential if we approach it thoughtfully. We don't have to wait for perfect systems. We can start small, learn and build the organizational capacity. Now, the real opportunity is not in the technology itself, but in how we use it to strengthen our ability to serve our communities, address inequities and respond more quickly to emerging threats. AI won't replace public health or its practitioners, but it can help us do public health better.

JOHNSON: 

If you'd like to read Guidehouse's State of GenAI Today report, we have a link right to it in the show notes. We'll also have a link to more information about ASTHO's Innovation Advisory Council.

That'll do it for today. Meet us right back here tomorrow morning with more ASTHO news and information. I'm Summer Johnson, you're listening to the award-winning Public Health Review Morning Edition. Have a great day.