The carrier is increasing build activity by 20% while integrating AI to streamline construction
Verizon is accelerating its network build activity by over 20% year-on-year, pushing past 30 million fiber premises passed with the Frontier acquisition, and betting heavily on AI and edge computing as the foundation for what comes after 5G. The carrier is also deploying AI internally to reduce friction in its construction workflows.
SVP and Chief Network Officer Srini Kalapala laid out the carrier’s priorities in a wide-ranging conversation with RCR Wireless’ Jeff Mucci at Connect (X), covering everything from virtualization and network APIs to satellite partnerships and quantum-era security. The gist is that Verizon is doubling down on mobility and broadband, evolving the cost structure without pushing the burden onto build partners, and positioning the network for a world where AI and autonomous applications need real-time compute at the edge.
Mobility and fiber network expansion
According to Kalapala, mobility and broadband remain Verizon’s two core service areas, and that’s where the bulk of capital is going. Kalapala said execution on build activity is up roughly 20% over last year, with new builds specifically also tracking above 20% year-on-year. That’s a meaningful step up, and it’s tied directly to growing customer usage and the need to deliver both capacity and coverage.
On the fiber side, Verizon is targeting 2 million premises passed this year. The Frontier acquisition pushes the company’s fiber footprint past 30 million premises passed, which is a substantial expansion of the consumer-facing fiber base. That same fiber does double duty — Verizon uses its consumer infrastructure to connect its own network assets across the country, which is part of why the carrier has historically leaned into being first in fiber. It launched its fiber business in 2003, and Kalapala was clear that he thinks Verizon knows how to deploy it at the right cost.
Hardware virtualization and edge computing
Verizon is arguably the most aggressive U.S. carrier on virtualization, and Kalapala didn’t shy away from saying so. The company has 40,000 sites running on virtualized distributed units (vDUs) — a number he claims no one else matches today. That’s on top of a fully virtualized back office, packet core, and IMS infrastructure.
The more interesting piece is Mobile Edge Computing. Verizon has MEC infrastructure deployed across 22 U.S. sites, putting cloud compute right next to the packet core so workloads don’t have to round-trip to a distant central data center. The latency benefits are obvious on paper, but the use cases have been slow to materialize — Kalapala acknowledged as much, noting that the autonomous driving and drone applications the industry expected to take off with 5G have taken their time arriving.
That said, he thinks the pieces are now coming together. His read is that edge will become significantly more important over the next decade, and the open question is how easy carriers can make edge infrastructure to deploy and consume.
AI integration and network APIs
AI inside the network itself is where Kalapala got most animated about the future. The vision is a move away from static service tiers and toward dynamic provisioning, where the network adapts in real time to whatever a user is actually doing.
Network APIs are the mechanism. The industry-wide goal is to let machines invoke an API to request specific characteristics, like low latency, high downlink, more uplink, or whatever else, for the duration they need it, without a human in the loop. It’s a shift in how network capabilities get packaged and sold, and it’s the kind of capability that only really makes sense once 5G standalone is broadly deployed and 6G is on the horizon. Kalapala framed AI and the networks as evolving together from here, which is probably the right way to think about it.
Partner collaboration and cost efficiency
When carrier CEOs talk about lower cost, the room full of tower companies and construction partners tends to assume they’re the ones who’ll pay for it. Kalapala’s answer for that, however, was that efficiency, in his framing, isn’t about pushing costs onto partners — it’s about removing friction between Verizon and the people who build for it.
Verizon’s teams produce architecture diagrams that then get converted into construction diagrams handed off to build partners. The fallout rate between the two has historically been high, and bad inputs lead to rework, wasted material, and application errors. AI tools are now being used to translate architecture diagrams into construction diagrams more accurately, cutting human error out of the handoff. Whether that fully translates into a healthier commercial relationship with the build community still remains to be seen, of course.
Satellite connectivity as a complementary tool
On satellite, Kalapala was measured. He called direct-to-device satellite the most exciting thing out there, but also offered a pretty pointed personal aside that the industry’s valuations are running ahead of what’s actually been delivered on the D2D side.
His position is that satellite is genuinely complementary to terrestrial networks, particularly for areas where fiber, cable, or fixed wireless are difficult to deploy. Verizon’s plan is to lean on partnerships rather than build its own satellite capability, which tracks with how the carrier has approached most adjacencies — find the right partner, integrate the service, keep the focus on the core.
Security in the quantum and AI era
Security came up at the end, and Kalapala’s framing was that AI cuts both ways. The defensive side has gotten harder, fast — if a malicious actor can use AI to generate a new threat in minutes, defenders have to close gaps and ship patches at the same speed.
But there’s a commercial angle too. Quantum-safe networks and secure AI-at-the-edge offerings are both things Verizon thinks it can package and sell to enterprise customers, particularly given the trust telecoms have built up around delivering secure connectivity. Whether that materializes into meaningful enterprise revenue is a longer conversation, but the opportunity is real, and Kalapala was clear he sees both sides — the urgency on defense, and the upside in shaping a new market.
Table of Contents
