Tech Explained: Here’s a simplified explanation of the latest technology update around Tech Explained: Billings, Mont., Implements Its First AI Security Policy in Simple Termsand what it means for users..
(TNS) — City council member Andrew Lindley wants Billings to refine its approach to technology.
During Monday night’s city council meeting, Lindley proposed the city create a Technology Advisory Commission. It would, he explained, be made up of industry members who could provide “extra guidance,” to the city on five key technological issues, including the city’s cybersecurity policies, the technology Billings uses to interact with citizens and AI governance within city government.
On at least one of those aspects, the city has already beaten Lindley to the punch. In April city staff under the direction of City Administrator Chris Kukulski and other senior officials implemented its first AI security policy, as part of an update to the city’s IT handbook.
“AI was becoming more and more significant, and it was something that we felt like an IT policy manual should have in it,” Billings IT Director Jeff Sprock said. “It’s embedded in a lot of the technology that we use and so we wanted to make sure that we had something where we were getting in front of it.”
In its current form, the policy addresses topics such as privacy, transparency and ethical usage.
“AI has the potential to enhance city operations, improve services for residents, and increase efficiency,” the policy’s first paragraph states. “Still, its use must be governed by strong security measures to protect data, prevent misuse, and maintain public trust.”
EVERYDAY USAGE AND DATA SECURITY
City staff are allowed to use AI to assist in basic everyday jobs tasks such as performing web searches or writing emails, Sprock and Kukulski both confirmed.
“We know that it’s being used because we can see the tools like Claude and ChatGPT being accessed by staff,” Kukulski said. “We look at it like Google or Microsoft Edge or Bing search tools.”
Kukulski said he has personally used AI to help him edit and polish his writing.
“My own use is I’ll draft a memo to the city council on a topic, and I’ll ask AI to simplify or improve it,” he said. “I don’t consider myself a great writer, so sometimes what I’ve found is it saves me a lot of time. Instead of spending an hour rewriting and rewriting something three times I can get all my ideas on paper and have it be cohesive quickly.”
While acknowledging its positive capabilities, Kukulski said he always double checked any of his work created with the assistance of AI and encouraged all city employees to do the same.
“You have to read and reread and double check anything that comes out of it because sometimes it hallucinates, sometimes it brings in things that just aren’t accurate,” he said. “So at the end of the day, we’re all responsible for what we produce.”
Both officials emphasized the security policy forbade the upload of sensitive personal information — including personal information, police materials and court materials — to AI models of any kind.
“We’ve been very vocal in saying you can’t put PII, (Personably Identifiable Information) into AI,” Sprock said. “You can’t put CJIS (Criminal Justice Information System) information into AI. Those things are not appropriate.”
In addition to what AI is being used for, the city also monitors and — when necessary — restricts the type of AI models being used by city staff.
Under Billings’ security policy, AI systems must meet industry standard data security principles, undergo a risk management assessment and receive proper clearance before implementation.
“The big concern with AI is you don’t know where the data necessarily is being stored,” Kukulski said. “It’s not concrete, so you can’t 100% say you know that that data is completely contained.”
Sprock pointed to DeepSeek, a Chinese owned and operated AI company, as an AI system the city banned for its questionable data practices.
“AI’s a new thing, and we’re grappling with how to approach it.” he said. “So, we’ve discussed it. We sat down with the city attorney and discussed it with her too.”
MAINTAINING PUBLIC TRUST THROUGH TRANSPARENCY
One of the biggest public debates about the ethical usage of AI is transparency regarding its usage. When does the usage of AI need to be disclosed versus when does not?
The city’s policy says it is committed to being transparent with the public about its AI usage and directs employees to “make efforts to explain AI decisions to the public, particularly in cases where AI influences critical decisions impacting city residents (e.g., law enforcement, resource allocation).”
Kukulski and Sprock both acknowledged AI transparency was particularly important for institutions with the power to impact large swaths of public and private life, such as local governments.
“If something is AI generated there needs to be transparency,” Kukulski said.
Further ensuring transparency, a reporting system for AI misuse was implemented as part of the broader city policy in April with punishments ranging from suspension of IT privileges up to termination and potential legal action for severe abuses.
AN ONGOING PROCESS
The AI industry is constantly growing and changing, making it especially difficult to regulate.
Despite this, Sprock said he and his team were always working to stay ahead of the AI curve and were already looking to make updates to the city’s AI security policy, even though it was only implemented less than a year ago.
“As quickly as it’s moving, it’s that you could have adopted something 90 days ago and already new information is coming out,” he said. “So, the fact ours was written 18 to 24 months ago and put in place in the past year. It’s the kind of thing that’s moving much, much quicker than we’re accustomed to.”
A full copy of the city’s AI Security Policy can be found inside its IT Policy Handbook accessible on the city’s website.
©2026 the Billings Gazette, Distributed by Tribune Content Agency, LLC.