AI could offer a helping hand for treasury management, study finds
da-kuk via Getty Images
More effective treasury management, including via the use of tech, can help local governments build financial resilience amid budget and staffing shortages, one expert says.
For states and localities navigating budget and economic uncertainty, treasury management frameworks that include efficient technology could help them build resiliency into their financial systems, according to a recent study conducted by the University of Chicago and commissioned by DebtBook, a financial management software provider.
In the last two years, relationships between the federal, state and local governments “have become strained,” said Justin Marlowe, co-author of the study and research professor at the University of Chicago Harris School of Public Policy, where he also serves as director of the Center for Municipal Finance.
“As a result of that, you now have state governments that have more uncertainty than ever about their budgets because they're getting squeezed on Medicaid, they're getting squeezed on unemployment benefits and they're getting squeezed on big infrastructure projects by the federal government — and that ultimately flows down to local governments,” he said, adding that these conditions could cause regular cash flows to be disrupted, impeding government operations and resources that rely on consistent budgeting sources.
That’s where technology solutions that enhance and modernize how municipal cash is collected, invested and spent plays a crucial role in states and localities’ financial outcomes, according to the study.
Indeed, it found that municipalities that had more formal treasury management practices — such as documented policies, board oversight and regular audits — in addition to tech that enabled more advanced financial data forecasting and analysis, had better financial outcomes, like higher credit and lower borrowing costs, Marlowe said.
In Wauwatosa, Wisconsin, tech modernization has helped the city rake in approximately $3 million annually through investment earnings, which make up 3% of the city’s general fund budget, according to the study. The city, for instance, transitioned its debt portfolio from spreadsheets to DebtBook, implemented a 24-month cash flow forecast and expanded electronic payment services for customers.
Such capabilities enable municipalities to track and allocate money more accurately and efficiently, which can be the difference between municipal leaders being able to fund full-time positions, library services and capital projects, Marlowe explained.
In particular, artificial intelligence is gaining traction among state and municipal governments looking to further streamline and optimize treasury management, he said.
“Emerging research shows that large language models can significantly improve cash flow forecasting accuracy for governments that lack codified forecasting methodologies, and machine learning approaches to anomaly detection offer new tools for fraud prevention,” the study states.
AI-enabled treasury management tools can also help governments better deal with additional challenges impacting their cash management capabilities, Marlowe said. For instance, 45% of treasury professionals in the public sector use Microsoft Excel as their primary cash management tool, and 29% said treasury operations were managed by only one or two full-time employees, according to the study.
Indeed, a 2024 study found that AI, especially tools like ChatGPT, could help small towns in Nebraska facing severe staffing and budget shortages with assessing their finances and projecting budget forecasts.
In Mt. Lebanon, Pennsylvania, local officials have seen invoice processing times reduce from weeks to days after adopting an AI-enabled management platform in 2022 that helps expedite and optimize electronic processing.
In Texas, the Schleicher County Treasurer's Office also saw efficiency gains after adopting generative AI-enabled software in 2025 that cut the time for staff to calculate the county’s equipment leases by two hours.