‘Don’t be afraid to try’ AI in public health, state official says

Maskot via Getty Images
Iowa and Oklahoma are seeing program efficiency and staff productivity gains as state agencies increasingly weave artificial intelligence into their operations.
Like many state agencies across the U.S., the Oklahoma Healthcare Authority is trying to balance a tightening budget with the need to continue serving residents with critical resources and services. Artificial intelligence is helping agency staff enhance productivity, save costs and resolve public problems, state officials recently told Route Fifty.
State leaders are laser-focused on delivering services like public health and social services that residents rely on, but they are also well aware of the roadblocks to do that efficiently, including legacy tech, budget constraints and low staff capacity, state officials said during a Future Nation webcast hosted by Route Fifty and Google Public Sector last week.
For example, the Oklahoma Healthcare Authority, which administers the federal Medicaid program in the state, “just had a really challenging budget cycle,” said Stephen Miller, chief technology officer of OHCA.
“We’re a $14 billion agency, and [the state] legislature decided there was about $250 million we didn’t need, so we had to find that $250 million,” Miller said. OHCA requested that lawmakers provide an extra $495 million in funds for fiscal 2027, half of which the Oklahoma Legislature approved earlier this year.
Such funds are critical for enabling the agency to implement systems and tools to manage beneficiaries’ applications, verify their identities and complete other administrative tasks, Miller said, adding that having “a very tight budget” means “we need savings from our technology.”
One way OHCA is addressing budget and capacity challenges through technology is with an AI tool that helps optimize documentation review. The agency is working with tech consulting firm Slalom to use AI to review documents, pay stubs and other resources to calculate people’s modified adjusted gross income as staff determine their eligibility for Medicaid, Miller explained.
“Instead of taking 20 minutes to review that [documentation], the AI will calculate the income and provide that back in one minute or less,” he said, adding that the tool has made “about a $6 million impact” in its first year of operation. Such efficiencies are especially crucial for state agencies like OHCA facing updated Medicaid rules from the federal government, including more stringent work requirements that go into effect Jan. 1, 2027, Miller added.
OHCA also launched an AI chatbot in April that was developed to assist staff in fielding resident questions about their health insurance coverage. The tool, which Google Public Sector helped build, aids OHCA’s call center staff in addressing resident inquiries regarding eligibility to, fielding approximately 95,123 voice calls and 308,834 since its rollout, which has reduced the staff’s tier one response by 70%, Miller said.
Such tech solutions and others could help OHCA save up to $70 million by automating and optimizing how agency staff are able to complete their work and connect residents to services quicker, according to Miller.
In Iowa, AI is enabling child welfare workers to more efficiently manage cases, improving the quality of service they deliver to residents and their work-life balance and morale, said Janée Harvey, director of the Division of Family Well-Being and Protection at the Iowa Department of Health and Human Services.
“What we see in the child welfare space is we have mission-driven social workers who are out there trying to do the good work with families, and yet they're spending most of their time sitting in a cubicle in front of a computer typing notes,” Harvey explained.
HHS is also working with Google to use AI to streamline and simplify case management among staff, which is particularly helpful when staff have to share cases with one another. With an intelligent case summary tool, AI is reviewing years worth of case notes to create summaries, including relevant information like recent family separations or a parent’s participation in a rehabilitation program to provide “consumable data that helps them in their social work practice,” Harvey said.
Performance data shows that “more than half of [staff’s] time is going to be given back to them,” she said. Harvey also pointed to an internal chatbot that supports staff as they work with families with complex case situations by suggesting applicable policies or procedures when prompted.
More broadly, as AI technology assumes administrative work like data entry and notetaking, HHS staff are “going to be better at their work, and they're gonna have more time to do their work,” Harvey said.
Ultimately, state agencies have “got to find savings and be able to … make a better environment for our members and workers and providers. That's what it's all about,” Miller said. Exploring and implementing innovative solutions like AI is a major way to do that, he added, “but you can’t be afraid to go try it.”




