Almost every broadband provider is asking the same question right now: how do we use AI without getting it wrong? At the 2026 NTCA Fall Conference, the session “Building Your AI Roadmap: A Practical Journey for Rural Broadband Operations” put that question to operators who are already working through it. Their answers were refreshingly practical: start with people, put guardrails in writing, and solve real problems rather than chasing the technology.
Start with people: education and honest conversations
Chuck Diesbeck, CEO of South Slope, described an education component as a foundation for his team, helping employees understand what AI actually is before asking them to use it. That groundwork matters because the concern is real. Longer-tenured employees tend to worry most that AI will take their jobs, and leaving that unaddressed makes adoption harder.
Diesbeck also pointed to the pressure of keeping up, both with the pace of change in AI itself and with competitors who are moving quickly. His advice for operators who feel behind is to have an open discussion, and to start at the top, in the boardroom.
Put policy and guardrails in place
South Slope has established an AI policy with guardrails, which is sound advice for all providers. Johnathan West, General Manager of Twin Lakes Telephone Cooperative Corporation, added a pointed piece of advice: don’t use the free versions of AI tools. Free tiers raise open questions about how company data is handled, which is exactly what a policy should address. West also emphasized encouraging employees to grow on their AI journey rather than simply restricting them.
Build a strategy that looks inside and out
Scott Alcott, CEO of Innovative Systems, argued that every provider needs a corporate strategy for AI that covers both internal and external use. He framed it around three questions every operator should be able to answer:
- How are your employees using AI?
- How are you governing it?
- How are your customers using it?
Telecom, he said, should ground AI in reality just as it does with everything else in the network, with solid systems of record and a clear policy framework behind it. Policy controls also need to reach the operational level. His illustration: you don’t want an automated process resetting a customer’s modem in the middle of a job interview. Good governance means thinking through the real-world consequences of automated actions, not only the technology.
Feeling behind? Solve a problem you already have
West’s advice for operators who feel late to the party was to resist the urge to create a problem for AI to solve. Instead, find a problem you already have and see whether AI can help. That keeps projects tied to business value and makes it easier to measure whether they worked.
How boards should think about AI governance
The panel also addressed the board’s role. The guidance was to avoid stovepipes. When AI tools and modules are deployed in isolation, each one is a blind spot; when they are governed together, each can inform the others. Boards that treat AI as a company-wide capability, rather than a series of departmental experiments, are better positioned to oversee it.
Where to begin
- Start the conversation at the top, including in the boardroom, and be honest about where you are.
- Educate your team on what AI is, and address job-security concerns directly.
- Write an AI policy with guardrails, and avoid free versions of AI tools.
- Pick one existing problem and test whether AI can help.
- Define your strategy for internal use, governance and customer use.
