Local governments are moving beyond broad discussions about artificial intelligence (AI) and beginning to make practical decisions about how it should be used, governed and supported.
Local officials, in a conversation that NLC convened at its Summer Board and Leadership Meeting, described how their communities are approaching AI from two connected but distinct directions. Inside city hall, leaders are creating governance structures, identifying operational challenges and testing targeted uses of AI. Outside city hall, they are confronting growing public interest in the data centers and other infrastructure needed to support an increasingly digital economy.
NLC brought together Cleveland, Ohio Council President Blaine Griffin; Avondale, Ariz. Councilmember Gloria Solorio; Golden, Colo. Mayor Laura Weinberg; and Pamela McKnight, Louisville, Ky.’s first chief AI officer, to share their experience with their peers.
Building Governance Before Expanding Use
Panelists stressed that local governments need to build internal structures for managing AI before they scale its use.
Cleveland established an Office of Urban AI through legislation, creating a formal home for the city’s work on emerging technology, data and digital services. The office oversees Cleveland’s open data portal and supported the relaunch of 311 as a central entry point for city services. The portal also includes a property insights tool that brings together information residents, businesses and city staff may need when evaluating a property.
Avondale began with an AI governance committee and a limited pilot involving approximately 14 employees using a general-purpose large language model. Rather than pursuing highly visible applications, the city focused on identifying recurring operational problems that AI might help address. Councilmember Solorio pointed to police report writing as one example. Preparing reports can consume significant officer time, and Avondale is testing whether AI can use information from body-worn camera footage to help officers produce initial drafts. Officers remain responsible for reviewing and approving the final report.
Councilmember Solorio’s advice to other local leaders is to start with defined problems, not with technology. A use case may sound innovative, but it has limited value if it does not address a genuine operational or resident need.
Louisville’s Portfolio of Pilots
Louisville took a citywide approach by creating a chief AI officer position and asking agencies to identify their most significant operational challenges before the city developed solicitations or selected tools. That process produced five initial pilots.
The most advanced focuses on permitting. Louisville is working with private sector partners to test tools that can help applicants identify the correct permit type and submit more complete applications for residential and small commercial projects. The pilot is being tested with a group of local developers and builders.
Other pilots focus on improving 311 services and internal AI governance, including the identification of “shadow AI,” or tools employees may be using without formal review. The goal is to understand current practices, identify gaps and establish appropriate safeguards. The remaining pilots involve public records and redaction, as well as computer vision and tools for the city’s real-time crime center. Louisville is also considering how they can support traffic management and technology-assisted translation services.
Data Centers Raise a Different Set of Questions
While cities are establishing processes for the use of AI within government, they are also learning how to talk with residents about the infrastructure associated with its growth.
Mayor Weinberg described a gap between the technical discussion surrounding data centers and the concerns residents bring to public meetings. Communities are being asked to consider projects with real implications for electricity, water, land use and utility costs. Officials who explain these tradeoffs clearly and early are finding it easier to conduct transparent engagement with residents.
Cleveland recently adopted a moratorium on new data center development to give the city time to evaluate its zoning and land-use rules. Council President Griffin described the pause as an effort to avoid repeating earlier periods of industrial growth when development moved faster than local policy. He also noted that cities want to stay economically competitive. He described Cleveland’s approach as taking enough time to get the rules right, without taking so long that they lose investment.
In Avondale, one project attracted limited attention during its initial review but drew packed public meetings once construction became visible. Local officials reported successful negotiations with the developer — including securing an agreement for support of a local grocery store — in an area that had a difficult time securing investment. That experience has led Avondale toward earlier and more proactive community engagement on future projects and clearly communicating the tradeoffs with residents, which is an approach other officials in the room said they were watching closely.
Looking Forward
What’s clear from Cleveland, Avondale, Golden and Louisville is that local governments succeed when they start with clarity: clear problems to solve when adopting AI, and clear benefits to communicate when proposing infrastructure like data centers. Both require the same groundwork, governance structures, defined goals and early resident engagement to build lasting public trust.
About the Author
