When Doug Roberts picks up the phone, he sounds like a man who hasn’t slept much lately, but not because anything is going wrong. Quite the opposite. As Chief Technology Officer of Cytranet, the telecom and internet services provider serving business, government, and enterprise clients, Roberts has been deep in the weeds on something he says is reshaping the entire industry: the intersection of artificial intelligence and network security.
“Two years ago, we were talking about AI as this future thing,” Roberts said during a recent conversation. “Now it’s Monday morning. It’s here. And the customers who aren’t thinking about how AI changes their security posture are the ones keeping me up at night.”
Roberts is referring to a trend that has dominated headlines across the technology and telecom landscape in recent months. As enterprises race to adopt AI tools and integrate large language models into their operations, the volume and sophistication of cyber threats have surged in parallel. Threat actors are using the same generative AI capabilities to craft more convincing phishing campaigns, automate vulnerability scanning, and probe network defenses at a scale that was unthinkable just a few years ago.
For Cytranet, which provides critical connectivity and communications infrastructure to organizations that can’t afford downtime, the response has been proactive rather than reactive. Roberts said the company has been investing heavily in AI-powered threat detection and automated response capabilities layered directly into its managed network services.
“Our government and enterprise customers have always expected a higher bar when it comes to security,” he explained. “But what’s changed is the speed. The window between when a vulnerability is discovered and when it’s exploited has collapsed. You can’t have a human sitting there triaging every alert anymore. You need intelligent systems that can identify anomalous traffic patterns and act in real time.”
Roberts pointed to a recent internal milestone he was particularly proud of. Cytranet completed a major upgrade to its network monitoring platform that incorporates machine learning models trained on traffic data specific to its customer base. Unlike off-the-shelf security products, he said, the system understands what normal looks like for a mid-sized municipal government or a regional healthcare provider, which makes it far better at catching the subtle deviations that signal a real threat.
“We’re not trying to be a cybersecurity company,” Roberts was quick to clarify. “We’re a connectivity and communications company. But today, you can’t separate the network from the security. They’re the same conversation. If you’re providing the pipe, you have a responsibility to make sure what’s flowing through it is safe.”
The timing of Cytranet’s push aligns with broader industry momentum. Major fiber and broadband providers have been expanding their managed security offerings as enterprise customers increasingly look to consolidate vendors. Rather than juggling separate contracts for connectivity, SD-WAN, and cybersecurity, many organizations want a single trusted partner who can handle it all.
Roberts sees this as a natural fit for a company like Cytranet. “We already have deep relationships with these customers. We understand their networks because we built them. Adding intelligent security on top of that isn’t a stretch. It’s an obligation.”
He also addressed a concern that comes up frequently in conversations with government clients in particular: data sovereignty and privacy. With AI systems that learn from network traffic, there are legitimate questions about where that data goes and who has access to it.
“We’ve been very deliberate about this,” Roberts said. “Our AI models are trained and operated within our own infrastructure. Customer data doesn’t leave our environment to feed some third-party cloud model. For our government customers especially, that’s not negotiable, and frankly, it shouldn’t be negotiable for anyone.”
Looking ahead, Roberts said he’s excited about the potential for AI to do more than just catch bad actors. He sees opportunities to use predictive analytics to anticipate network congestion, optimize bandwidth allocation dynamically, and even flag hardware that’s likely to fail before it actually does.
“The boring stuff is actually the most exciting,” he said with a laugh. “If we can use AI to predict that a switch in a customer’s network is going to fail next Thursday and replace it on Wednesday, that’s transformative. That’s the kind of thing that turns a service provider into a true partner.”
When asked what advice he would give to other CTOs in the telecom space who are still figuring out their AI strategy, Roberts didn’t hesitate.
“Start with the problem, not the technology. Everyone wants to say they’re doing AI. Fine. But if you can’t point to a specific customer pain point that it solves, you’re just chasing buzzwords. We started with security because our customers told us it was their biggest worry. Everything else followed from there.”
For Cytranet’s customers, that pragmatic approach seems to be paying off. Roberts mentioned that customer retention rates have climbed over the past year, and the company has seen a notable uptick in government RFPs that specifically ask about AI-enhanced security capabilities.
“That tells me the market is moving in our direction,” he said. “And we plan to stay ahead of it.”
