A local leader's guide to using generative AI (without losing your voice)

If you work in local government or community organizing, you know the drill. Your inbox is overflowing with constituent emails. There's always another council meeting to prep for, another presentation to a neighborhood group, another newsletter that needs writing. And social media—there's always more social media.

We've written before about how to show up authentically on social media as a public official and why tools like Voxifly can help you stay consistent without burning out. But let's zoom out for a moment and talk about everything else you write and communicate. How can tools like ChatGPT and other generative AI systems support your work without sacrificing the trust, accuracy, and local voice that makes you effective?

Think of this as a practical manual for local elected officials, community leaders, and communications staff who want to use AI responsibly in their day-to-day work—without sounding like a robot wrote their press releases.

Before you start: setting some ground rules

The thing about AI tools is that they're incredibly helpful, but only if you establish some boundaries first. These don't need to be formal policies at the start—just a few rules of thumb that everyone on your team understands.

The most important question to ask yourself is: what shouldn't go into an AI tool? A good rule is simple: don't paste anything that would make you uncomfortable if it showed up in tomorrow's newspaper. That means no sensitive constituent details, like names paired with specific problems, health information, or immigration status. It means no internal HR matters—personnel issues, performance reviews, complaints. And it definitely means no non-public negotiations, whether that's early-stage development projects, contracts, or legal strategy.

Now, you can still work with these topics. You just need to anonymize and generalize. Instead of asking an AI to "rewrite this email from Jane Smith at 123 Oak Street about her code violation appeal," try something like "help me rewrite an email responding to a resident who recently received a code violation and is worried about the fine. Keep it empathetic, clear, and firm about the process." See the difference? Same goal, but you're not feeding someone's personal information into a system you don't control.

The second rule is equally important: AI output is always a draft, never the final step. Your staff can absolutely use AI to generate ideas, outlines, or first drafts. But a human needs to be responsible for checking facts—especially dates, dollar amounts, and laws—adjusting tone, adding local detail, and making sure everything aligns with your values and existing policies. This matters especially for local leaders whose work is deeply relational. Your community isn't looking for perfectly polished copy. They're looking for signs there's a real person on the other end.

One more thing that will make your life easier: capture your "voice guide" somewhere you can reuse it. AI is much more helpful when it knows how you sound. Spend a few minutes writing down who you are (mayor, council member, nonprofit director), who you're talking to (residents, volunteers, local businesses), how you want to sound (approachable, plain-language, hopeful), and what you avoid (jargon, partisan language, insider acronyms). Something like: "I'm a city council member in a small town. My communication style is friendly, calm, and practical. I avoid partisan language and focus on solving problems with my neighbors. I prefer plain language at about an 8th-grade reading level." Keep that handy, and you can paste it into prompts whenever you need AI to sound like you.

Making everyday writing faster and clearer

Let's be honest: you probably spend a good chunk of your day in your inbox. Generative AI really shines at turning rough ideas into clear, respectful messages—and that can save you hours every week.

Take constituent emails, for example. You don't want AI answering on your behalf unsupervised, but you can absolutely ask it to help with drafts. Let's say you're responding to a resident who's frustrated about a new parking policy near the downtown business district. You might prompt the AI: "I'm a county commissioner responding to a resident who is frustrated about a new parking policy near the downtown business district. Please draft a short, respectful email reply that acknowledges their frustration, briefly explains why the policy changed, and invites them to a public meeting where we'll review feedback. Keep it under 250 words, in plain, friendly language."

From there, you edit. You add specific local details—street names, dates, locations. You make sure the explanation matches your actual policy decisions. You adjust any phrasing that doesn't quite sound like you. The AI gave you a solid starting point, but you're still the one who made it real.

AI is also great at stripping out jargon, which tends to accumulate in local government communication like dust on a shelf. You can paste in a paragraph from a staff report and ask it to rewrite the text so it's clear to a busy parent who doesn't work in government, avoiding acronyms and using plain language around a seventh-grade reading level. You can ask it to shorten something to three bullet points, translate it into Spanish while keeping a respectful tone, or turn it into a text message you could send to meeting registrants.

And if you're drowning in long documents—staff reports, consultant memos, grant guidelines—AI can give you a quick overview. Ask it to summarize the key points for a city council member who has five minutes to prepare before a meeting, focusing on the decision they need to make, any budget implications, and potential community impacts. Just always keep the original nearby. For local leaders, this can make the difference between walking into a meeting overwhelmed versus prepared to ask good questions.

Turning complex issues into accessible explanations

Local governments work on issues that are both technical and close to home: zoning, stormwater, road work, school funding, utility rates. AI can help you explain these clearly to the people who live with the outcomes.

Say you're replacing aging water lines on Maple Avenue, and it's going to affect water service for about 200 households for several hours. You might ask AI: "I'm a communications specialist for a small city. We're replacing aging water lines on Maple Avenue, and it will affect water service for about 200 households for several hours. Write an explanation for residents that briefly explains why the work is necessary, sets expectations about noise, traffic, and water shutoffs, emphasizes safety and long-term benefits, and invites residents to contact us if they have accessibility concerns. Make it friendly and straightforward."

What you get back can become a webpage or FAQ, a printed handout for doors, or a script for staff answering phone calls. The key is that you're still the one providing the local context—AI just helps with structure and readability.

And here's something to watch out for: AI doesn't know your neighborhoods. It will happily make up street names, landmarks, or organizations if you don't provide them. So give it the real names of parks, schools, and neighborhoods, or ask it to leave placeholders where you'll add specifics later, like "[insert community center name]." Always double-check any specifics in the output before publishing.

Preparing for meetings and presentations

Generative AI can be a surprisingly good rehearsal partner when you're getting ready to speak with the public. If you have a staff report or slide deck, you can ask AI to turn it into speaking notes. Give it a staff report about a proposed sidewalk expansion project and ask for three to five key talking points for a city council member to use during a public meeting, plus a two or three sentence explanation they can use if a reporter asks "Why are you prioritizing this project now?" Keep the tone straightforward, factual, and non-partisan.

You can then personalize those talking points—emphasizing the ones that matter most to you, adding stories or examples from your own experience, and using phrases your community responds well to.

You can also ask AI to role-play skeptical residents or stakeholders. Try something like: "Act as a resident who is skeptical about a new property tax levy that will fund park improvements in our county. Ask five to seven tough but fair questions about how we've managed funds in the past, whether this will raise costs for seniors on fixed incomes, and why we're investing in parks instead of roads. After listing the questions, suggest clear, empathetic answers a county commissioner could give, but keep them as bullet-point notes rather than full speeches."

This kind of practice can help you prepare for public comment, Q&A sessions, and media interviews in a lower-stakes environment where you can refine your thinking before the real thing.

Extending what you're already doing on social media

We've already dug into why public officials should embrace social media and how personality-driven content can build trust and engagement. AI can support what you're already doing in some practical ways: repurposing content across channels, creating versions of the same message tailored to different audiences (residents, local businesses, nonprofits, volunteers), and improving accessibility by generating alt-text for images or translating posts into commonly spoken languages in your community.

For instance, you might take a long council recap and ask AI to create one Facebook post that's conversational and friendly, one LinkedIn post that's more formal and focused on regional partners, and a short text message version under 160 characters that you could use with an SMS tool. You can also feed AI past posts that performed well and ask what patterns it sees in tone, structure, and topics, then have it suggest new post ideas that fit that style.

Watching out for common pitfalls

Generative AI is powerful, but it has well-known blind spots. The first is what people call "hallucinations"—confidently wrong answers. AI tools can invent facts, especially about local details. Incorrect dates for past meetings, nonexistent local ordinances, made-up statistics about your town or county. Protect yourself by asking it to separate facts from suggestions ("flag anything you are not certain about") and always going back to primary sources—your own records, staff, legal counsel—when it comes to laws, budgets, and official decisions.

Then there's the generic "AI voice" problem. Your community can usually tell when something sounds like canned copy. If all your messages start to sound like a corporate press release, people stop reading. Combat this by adding local detail (like "the playground at Riverside Park"), keeping your usual small quirks of phrasing, and occasionally sharing personal observations ("I was at last Saturday's clean-up and heard several neighbors mention..."). AI is a starting point, not the end product.

Equity and representation matter too. Because AI tools are trained on broad internet data, they can reflect existing biases. That's especially important when you're talking about historically marginalized neighborhoods, describing groups of people—youth, people experiencing homelessness, immigrants—or choosing which stories and images you highlight. You can ask AI to help you audit your own drafts: "Check this announcement for language that could feel stigmatizing or biased toward any group. Suggest alternative phrasing that is more neutral, respectful, and people-first." And then apply your own judgment and knowledge of your community.

Finally, don't forget about public records and transparency. In many jurisdictions, public officials' communications are subject to public records laws. If you're using AI to draft or summarize messages, treat AI-assisted drafts as part of the record when appropriate, keep copies of the final text in your usual archiving systems, and avoid relying on AI tools that don't allow you to export or save important content elsewhere. If in doubt, check with your city or county attorney or records officer.

Starting small with a simple workflow

You don't need a giant "AI strategy" to get value from these tools. Try this simple workflow for a week: pick one or two recurring tasks, like responding to constituent emails, drafting meeting recaps, or writing event announcements. Create a reusable prompt plus your voice guide and save it somewhere staff can copy and paste. Require a short human edit step—check facts, add local specifics, make sure it "sounds like us." Then reflect at the end of the week. Did it save time? Did it improve clarity? Were there any close calls with accuracy or tone? From there, you can decide where to expand and where you'd rather keep things fully human.

Bringing it back to the local level

The promise of generative AI for local leaders isn't about flashy technology. It's about getting more time back for the parts of your job that require judgment, empathy, and presence. It's about making your communication clearer and more accessible to real people, not just policy insiders. And it's about supporting small teams that carry a lot of responsibility for how your community understands what's going on.

Used thoughtfully, AI can help you do all of that without replacing your voice or your values. If you'd like to explore how tools like VoxiFly, GovFeeds, and AI-assisted workflows can fit into your office's day-to-day communications—across email, meetings, and social channels—we're always happy to talk.