Most of the AI conversation happens at the federal level or in big cities with dedicated innovation teams. Meanwhile, the office that actually answers the phone when a resident calls about their water bill is often a staff of two or three people covering finance, permitting, and customer service all at once. For those teams, AI isn't a strategy document — it's a practical question that has started arriving on its own: residents are using it, employees are curious about it, and expectations are shifting either way. The question most municipal leaders are asking isn't "should we have an AI initiative?" It's quieter and more grounded: what does this actually mean for how our office works day to day?

The landscape is shifting, whether or not anyone plans for it

A few forces are converging on local government at once. Residents who use AI assistants in their daily lives are starting to expect faster, clearer answers from their city or district. Staff are quietly experimenting with AI tools on their own — sometimes helpfully, sometimes in ways that raise records-retention and confidentiality questions nobody has answered yet. And the long-running workforce squeeze (see our piece on the retirement cliff) means every hour of repetitive work matters more than it used to. Together these mean AI is becoming part of the operating environment for local government — less a project to launch than a change in the weather to plan around.

Where the interest actually is

The research says municipal leaders are approaching this practically, not chasing hype. In ICMA's survey of municipal leaders, 67% named improving internal efficiency or automation a top AI priority for the next 12 to 18 months, with citizen services, infrastructure management, and compliance automation identified as the leading use cases. That framing matters: agencies aren't asking what's flashy, they're asking what could ease the load on staff who are already stretched thin — answering routine questions, summarizing long documents, drafting first passes of routine writing, finding the right policy in a filing cabinet's worth of PDFs.

The honest challenges

The same research is refreshingly candid about what's in the way. 63% of municipal respondents cited limited internal expertise or staffing as the biggest barrier to adopting or scaling AI, with funding limitations (41%), legacy systems (35%), and uncertainty around ethical or legal implications (52%) all shaping how fast agencies can move. Smaller communities face an extra gap: most of the guidance being written is aimed at big cities, while the more than 16,000 municipalities under 5,000 residents are largely left to figure it out on their own. And local government carries obligations most industries don't — public records laws, procurement rules, and the simple fact that residents can't take their business elsewhere — which rightly makes agencies slower and more deliberate than the private sector.

The questions worth asking before any tool

The municipalities navigating this well tend to start with policy and posture, not products. A few questions that come up in nearly every board room and department meeting:

  • Where is repetitive, well-documented work eating the most staff time — and would saving those hours actually matter to residents?
  • What's our position on staff using AI tools with public records, resident data, or draft documents? Who decides, and where is that written down?
  • How do we keep a person accountable for anything AI touches — especially anything that goes out the door to a resident, a council, or an auditor?
  • Is our underlying data — financial records, billing, personnel — organized and reliable enough that any future tool would have something trustworthy to work with?

That last question is easy to overlook and probably matters most. AI is only as useful as the information underneath it, and agencies still reconciling spreadsheets by hand or hunting through inboxes for the authoritative version of a number aren't ready to hand any of it to an assistant — human or artificial.

What this means for your office

Nobody needs an "AI strategy" to start. The realistic path is smaller: name one or two repetitive tasks worth studying, write down a simple policy for how staff may and may not use AI tools, and keep a person in the loop on anything that matters. None of this replaces professional judgment — and none of it should be pointed at anything higher-stakes than what your team is comfortable reviewing before it goes out the door. What thoughtful adoption can do is buy back a few hours a week for the people already doing the real work. For a two- or three-person office, that is not a small thing.

And whatever role AI eventually plays in your operations, the preparation looks the same: clean, well-organized systems and data. That work pays for itself today — and it's what will make anything you try tomorrow actually reliable.