Pottawattamie County Adopts Three-Tier AI Policy for Staff

Pottawattamie County officials have formalized guardrails around how county employees may utilize artificial intelligence, particularly when handling sensitive government information. The Pottawattamie County Board of Supervisors approved the policy this week, marking one of the more structured local government approaches to AI governance in the Midwest.

David Bayer, the county's chief information officer, explained that discussions about an AI policy began last year as the technology became ubiquitous in daily operations. "AI is definitely influencing what we read and see every day," Bayer said. "…It's everywhere."

The policy organizes county AI use into three distinct tiers. Tier one covers basic, non-sensitive uses—what Bayer described as "glorified Google search" scenarios where no sensitive data is involved and unlicensed AI tools may be used. Tier two involves county-licensed AI software that the IT department distributes for more advanced tasks. Tier three represents highly controlled AI systems handling sensitive data with strict oversight.

All county employees must complete both in-person and online training on compliance. The curriculum covers not just policy rules but practical application—how to use AI tools, what to look for, and how to document AI usage. "Not just the training on the policy, but also training on how to use AI, what to look for…how do I report or document that I'm using it. All those things will be part of that training," Bayer said.

The policy explicitly prohibits employees from inputting sensitive data into public AI platforms like ChatGPT. This includes criminal justice information, Social Security numbers, names, addresses, and phone numbers. The physical reality of this restriction means county workers can no longer copy-paste case files into a chatbot to summarize them—a workflow that would have been tempting but now carries compliance risk.

Enforcement relies on peer reporting rather than automated monitoring. Bayer's department is creating a system where employees can notify IT if they suspect AI misuse within county offices. "There's not going to be a lot of monitoring, other than we're going to have to rely on each another to know when someone's misusing it," Bayer said. (This trust-based approach will either work beautifully or collapse under pressure, depending on office culture.)

Independent reporting from Citizen Portal corroborates the three-tier structure and training mandates. The outlet confirms the board voted to adopt the policy following discussion about enforcement mechanisms and human oversight requirements.

The county's AI committee is simultaneously identifying ways AI could improve operational efficiency. This dual approach—restricting risky use while exploring beneficial applications—reflects the broader tension local governments face with emerging technology. The committee's work will determine whether AI becomes a net efficiency gain or just another compliance burden.

According to the Radio Iowa report, the policy addresses the kind of AI county employees would actually use rather than theoretical scenarios. This practical focus distinguishes it from more abstract corporate AI policies that often fail in implementation.

What makes this policy notable is its acknowledgment that AI use will happen regardless of policy. The question isn't whether employees will use AI tools—it's whether they'll use them safely. The tiered system accepts this reality while attempting to channel usage into approved pathways.

Other Iowa counties may watch this implementation closely. If the policy proves workable, it could serve as a template for smaller jurisdictions lacking dedicated IT leadership. If it creates friction or confusion, it may demonstrate why local governments struggle with rapid technology adoption.

The real test comes in execution. Training completion rates, reporting system usage, and actual compliance will determine whether this policy achieves its stated goals or becomes another document gathering digital dust. Whether county employees actually follow these guidelines remains the real question.

Artūras Malašauskas is an AI Systems Integrator with 20+ years of production-grade web engineering experience. He has designed, shipped, and scaled enterprise Python/PHP systems for logistics, SaaS, and public-sector clients. For the past year, he has focused exclusively on AI integrations: deploying open-source LLMs, building generative media pipelines (image, audio, video), and engineering multi-agent workflows for real production environments. His standard: reproducibility, security, cost-efficient inference—no vaporware. He documents and evaluates emerging AI tooling, separating verified capabilities from marketing noise. Technical editor at: muza-ai.eu, ai-verslas.lt, ai-naujinos.lt Connect on LinkedIn