- A cross-functional AI Center of Excellence fights skepticism and curbs shadow IT
- The strategy applies AI in, for, and by the telco to improve customer experience and drive revenue
- Early deployments of Copilot, ServiceNow, Qodo and Abnormal saved significant hours and blocked attacks
Overcoming internal resistance to AI has been one of the biggest barriers to implementation for C Spire, a regional U.S. service provider.
"Getting over the stigma has been one of our biggest challenges," Alan Jones, C Spire chief network officer, told Fierce Network. The general population is skeptical about AI, he explained, and people are just resistant to change. "You may have a guy that's been doing the same thing for 20 years, who's stuck in their ways. But once you get over the stigma, AI is a tool in the toolkit."
It seems odd to discuss AI stigma as investors pour hundreds of billions of dollars into AI, and telcos and their suppliers spotlight AI in their products and marketing. But AI skepticism is real.
Fewer than a third of Americans (31%) trust businesses to use AI responsibly, according to a Gallup report released last month. That's actually up from 21% in 2023. Among workers, 81% say they don’t use AI at work, and among those who do, most don’t consider it particularly helpful, according to a February Pew study. Meanwhile, a July study from MIT notoriously found 95% of organizations implementing AI get zero return on investment, despite $30-40 billion in enterprise investment.
Combating AI skepticism
The MIT study said the failures are less about technology capability than about implementation, and the study was greeted by skepticism. For example, Paul Roetzer, founder and CEO of the Marketing AI Institute, noted that the study had a narrow definition of success — financial return on investment within six months.
But C Spire sees value in AI, and overcoming AI skepticism is one of the chief reasons that C Spire stood up a Center of Excellence (CoE) to drive AI adoption within the company.
"We need to be forward-thinking, making sure to communicate to employees why we are using AI," Jones said. Developers need to go beyond just deploying tools. "Let users know what the tools are capable of and how they can make their day-to-day better, and then your adoption problem takes care of itself and the roadblocks start taking care of themselves."
C Spire built its AI implementation team with people who are passionate and already spending their own time learning to use AI. C Spire changed those people's job responsibilities to suit their passions, assigning them to spend half their time developing AI. "We had champions, and those champions started driving home the need for AI throughout the company," Jones said.
The Center of Excellence is a small group, which includes CEO Suzy Hays and a cross-functional team including networking, IT, marketing, operations and business service. The team serves as a steering committee for AI, deciding how to evaluate AI as a product that C Spire both uses and produces.
Avoiding 'shadow IT'
"There is no shortage of companies that will sell you something that says 'AI' on the label," Jones said. "Having a CoE can help evaluate and steer through some of that noise.
An additional important purpose of the CoE is to avoid "shadow IT," where line-of-business workers deploy AI resources without central coordination and oversight. This is something telcos and other businesses saw in the early days of the cloud, when users put cloud resources on their own credit cards; it can result in internal chaos and wasteful redundancies.
C Spire divides its use of AI into three areas: Implementing efficiency gains on the network; providing AI consulting and services and hosting AI applications in C Spire's data centers for business customers and hyperscalers; and using AI for internal end-user and customer service.
The service provider is using both ServiceNow AI and Microsoft Copilot extensively. It is using AI tools to track customer service level agreements, to ensure the company is meeting its commitments, as well as product and market research, financial analysis, best practices and industry benchmarks.
An internal ServiceNow AI tool allows users to track specific support cases, to assist in root cause analysis, providing case timelines and summaries.
Its strategy gets results. Over the last 30 days, using Microsoft Copilot, C Spire has saved 1,500 hours of time for a team of about 400 people, resulting in $65,000 monthly savings. Using its ServiceNow agent, CiSpire has saved over 346 hours for a team about the same size. The company uses Qodo (formerly Codium) for coding, saving about 95 hours in a month for a team of about 50 people. And Abnormal, an AI-based email protection tool, has caught nearly 2,000 phishing attempts aimed at users in C Spire, Jones said.
But the greatest benefit for AI is being able to reduce team members' workloads to free them to do more strategic work, he said.
C Spire is not alone in using a CoE to deploy AI strategically. Globe Telecom, for example, built a common AI infrastructure throughout the company to avoid wasteful redundancy.
"We have been building bots and agents so we can have our own internal tools that are unique to us," Jones said. The process has matured so that developers can work quickly. "If we don't have an agent for something right now, by the end of the day if someone asks for one, we will."
Prioritizing customer experience
C Spire is a privately held regional service provider serving customers in Mississippi and parts of Alabama and Tennessee with more than 22,000 miles of fiber. It provides home fiber internet, wireless and business IT solutions. It goes up against much larger players — Verizon, T-Mobile and AT&T. C Spire competes based on on-site, localized customized support and a 5G network with the fastest and most responsive download speeds in Mississippi, according to Ookla, C Spire says.
C Spire's goal in deploying AI — as with all its work — is to improve customer experience. "We're making sure we use AI in a way that makes the customer experience better for the customer rather than making sure we're using AI in a way that makes the customer experience better for us," Jones said.
For example, when deploying chatbots, the goal is not to speed up efficiency for the call center. "The goal is to make sure the customers are able to get the responses they were looking for in the way they were looking for them faster," he said.
Likewise, for network operation tools, the goal is not to save time for engineers; the goal is to make sure the mean time to repair (MTR) is reduced for the customer.
C Spire built an agent for supporting end-users within the company, to assist with routine tasks such as installing applications, providing guidance in using internal software and explaining company policy for cybersecurity. Previously, that work would have been done by IT staff.
"We don't have to tie up a valuable, skilled resource to answer a question that doesn't really pertain to them, or to walk through something that the end user could walk through themselves — they just need a guide," Jones concluded.
AI in, for and by the telco
C Spire's approach reflects the three common ways AI is applied by telcos. There's "AI in the telco," when telcos use AI to optimize the network; "AI for the telco," using AI for internal functions like customer service, billing and software development; and "AI by the telco," where telcos offer AI as a service (AIaaS). We talked about those themes at our recent Fierce Network Research live event in Dallas, where we discussed AI use cases by 3 Hong Kong, Globe Telecom and Orange Business.
Catch our recap of the event here.