Public Health, Meet AI: Inside the Second Public Health & AI Summer School
At the college’s 2026 Public Health and AI Summer School, 75 participants spent four days in learning to put artificial intelligence to work for population health — and built 20 working prototypes along the way.
Across four days this June, in the Grand Challenges Research Building on the University of Arizona campus, the Public Health & AI Summer School organized by the Zuckerman College of Public Health trained a diverse range of students to use AI tools for public health goals.
The central question for the AI Summer School was not whether Artificial Intelligence (AI) belongs in the public health field, but how public health professionals can use AI most effectively to protect population health.
This is the second year for the School, which ran from June 8–11, 2026, and drew 75 participants that ranged from epidemiologists and program manager to faculty and agency leaders, graduate students, and front-line practitioners. Participants came from across Arizona and from Colorado, Nevada, California, New York and Kansas, as well as Brazil. They represented state, county, and tribal health agencies. Two-thirds of the participants were women at a time when women are still underrepresented in the AI field.
What held this diverse group together was not a shared technical background. It was a shared starting point, the foundation of the program: public health first, AI second.
“We are not training engineers. We are training public health professionals to think clearly about AI — to use it well, question it honestly, and keep population health at the center. When you start from the problem instead of the tool, a room full of very different people can all find their way in.”
A workforce-first approach
Rather than teach AI as a subject for engineers, the Summer School treated it as a competency for public health professionals. Participants gained the ability to spot where AI fits a problem, to evaluate a tool with a critical eye, to talk about it clearly with colleagues and decision-makers, and to help build something that works. The design assumed no one needed to become a programmer to use these tools responsibly. They needed judgment, and a place to practice it.
Participants were able to choose between two in-person tracks. The AI Literacy track was built for people starting at the beginning. The AI Fluency track was designed for those with more experience who wanted to push further. An additional online track carried the program beyond campus for those who could not travel to Tucson. None of it was rigid — participants could move between tracks as their confidence grew.
The AI Maker Space
The heart of the curriculum was the AI Maker Space, a hands-on studio where teams formed around a real public health problem and worked it all the way to a demo-ready prototype. Projects ran along three streams: AI agents, data projects, and functional web applications. Teams picked the stream that matched their problem, not their comfort level, and spent the week moving from a rough idea to something they could show by the end.
By the final day, the teams had produced 20 working prototypes. One group built an Ebola surveillance concept. Another made a bilingual public health education site, aimed at communities that too often get information last and in a second language. A third put together a performance dashboard for mobile health units, the kind of tool a program manager could use to see where services are reaching people and where they are falling short. The range of prototypes said something about the people in the room: the same set of tools, pointed at very different problems, by people who knew those problems first-hand. On the last day the teams selected the top 10 projects to be presented to the entire group, and the top 3 projects were awarded scholarships for the online track.
A cross-campus effort
The program’s 15 speakers were not drawn from one college or department but from across the University of Arizona. Five U of A colleges and institutes took part — the College of Public Health, the College of Medicine–Tucson, the College of Science, the College of Agriculture, Life & Environmental Sciences, and the BIO5 Institute — alongside partners from outside the university, including Northeastern University, the City of Tucson and the Pima County Health Department. That wide array of speakers was deliberate. AI in public health does not sit neatly inside one discipline, and the people teaching it came from medicine, data science, agriculture and local government as much as from public health itself. Practitioners from city and county agencies kept the conversation close to the work that real health departments do every day.
Measurable gains in confidence
Surveys conducted both before and after the program told a consistent story. Confidence rose across every applied skill the program set out to build, skills such as explaining AI to a colleague, evaluating a tool, holding a strategic conversation about AI tools, and recognizing where AI could be useful in the first place. Self-rated knowledge of AI in public health climbed from 4.7 to 7.6 on a 10-point scale.
One result was less obvious but more telling. Participants’ self-rated technical comfort held steady, and in some cases dipped a little. Organizers read this not as lost learning but as honest recalibration — the natural result of seeing, up close, how much real AI work involves. People walked in with a vague sense of what these tools could do and walked out with a clearer, more grounded picture, which sometimes meant rating themselves more modestly than before. That is usually what understanding looks like.
Satisfaction was high, with a mean rating of 4.6 out of 5, and 97% of respondents said the program met or exceeded their expectations. A majority attended on employer funding, which indicates that institutions are starting to treat AI capability not as an extra skill but as a core competency for the job.
Participants who met the attendance and engagement requirements earned the Applied AI in Public Health micro-credential, a verifiable digital badge that turns four days of work into something portable they can put in front of an employer or add to a social media profile.
“The most rewarding moment is the last day, when teams present something they built themselves,” said Dr. Leal Neto. “Four days earlier, many of them weren’t sure AI was for them. That shift — from ‘this isn’t for me’ to ‘here’s what I made’ — is exactly what we set out to create.”
Looking ahead
The Public Health & AI Summer School is part of the college’s AI for Public Health Initiative, a multi-faceted project designed to deploy AI and AI training across our education and research programs. The college is determined to integrate the same competency-based teaching approach to AI tools into our graduate coursework and a growing set of programs.
Planning for a 2027 edition of the Summer School is already underway, this time in Phoenix under the banner PHX, bringing the program to the state's largest population center and a wider pool of agencies and practitioners. The name of next year’s Summer School points to our focus: PHX is Public Health X, where X stands for the challenges that AI can be applied to solve.
“I’m so pleased and proud of the way Onicio has brought together expertise from across the university to train our students, our faculty, and working public health professionals in the use of AI tolls to solve public problems. Public health is more important than ever, and we must master the use of AI meet to meet emerging health challenges.”
Learn more about the AI for Public Health Initiative at publichealth.arizona.edu/ai.
