Picture a resident with a simple question. Is the park open today? Is the water safe to drink? How do I apply for a permit? A few years ago, they would have searched, scanned a list of links, and clicked through to the official page.
Today, a growing number of people ask an AI assistant instead and the assistant gives them one answer. Here is the question every public information officer should be sitting with: when AI answers a question on the agency's behalf, is it using *the agency's* information — or someone else's?
The uncomfortable part
For most agencies, the honest answer is "someone else's." AI assistants tend to overlook the very places government communicators work hardest. The carefully maintained service page, the daily social posts; much of it is built in a way these tools quietly skip over. And when AI can't find a clear official answer, it does not go silent. It fills the gap with whatever is clear: an old news story, a third-party site, or a thread of residents guessing among themselves. So the answer a resident hears about a public service may not be wrong by accident. It may simply be the most *findable* version, and right now, that version rarely belongs to the agency.
Why this can't wait
This is not a someday problem. It is happening on every phone, today, whether or not any agency has a plan for it. And the cost compounds: the longer an unofficial source is the clearest one, the more it hardens into the default answer. The communicators who address this early become the trusted voice in their community. The ones who wait will spend the next few years correcting a record they could have owned from the start. The good news is that the fix is not technical. It needs no vendor, no budget, and no new platform, and it starts with content the agency already has.
What to do about it
Our new learn along and online course is a practical introduction to Generative Engine Optimization (GEO); what it is, where to apply it across the content an agency already has, and how to prioritize strategically so the official answer is the one residents get.
It comes in two parts that work together;
- A live learn-along webinar explaining how AI decides what to surface, and where to focus first.
- A self-paced online course with the full method, a worksheet, a self-check, a starter list of the questions residents actually ask AI, and the session recording
It is free for PIO Toolkit members. Non-members can join a session for $89, which includes the course.
👉 Wednesday, July 22, 12:00–1:00 PM ET.