Verizon is positioning its fiber network and real estate as infrastructure for the next phase of AI demand. The company says a dark fiber agreement worth more than $1 billion with Google is only one part of a wider plan to sell connectivity and edge computing capacity to AI-focused customers.
A $1 Billion Fiber Deal With Google
During an earnings conference call covering the second quarter of 2026, Verizon publicly described a partnership that uses Verizon dark fiber routes to connect Google data centers. The company presented the deal as part of a new business initiative called “AI Connect.”
The value of the Google arrangement is more than $1 billion. Verizon also expects to announce additional AI-related agreements by the end of the year. Dan Schulman, CEO of Verizon Communications, said those future deals would be “worth multiple billions of dollars in revenue over the next several years.”
The basic business case is straightforward: AI infrastructure needs fast links between data centers. Large technology companies are spending heavily on compute, and Verizon is arguing that its network can become a larger part of that spending cycle.
Schulman described the shift this way: “Now it really is about, ‘how do you connect data center to data center to assure that you can optimize to maximize the ever-exploding need for compute?’”
Why Dark Fiber Matters for AI Infrastructure
Dark fiber is unused fiber capacity that can be sold or leased for dedicated connectivity. In Verizon’s case, the company is looking at unused routes as a way to serve hyperscalers’ AI data centers.
The source article identifies Google as the first disclosed customer in this new push. Verizon’s expectation is broader than one customer, however. AI Connect is being positioned as a repeatable business line, not a one-off transaction.
For Verizon, the timing is tied to its larger fiber footprint. The company finalized a $20 billion acquisition of Frontier Communications at the start of 2026. That deal expanded Verizon’s fiber footprint from 15 states to 31 states along with Washington, DC.
That larger network gives Verizon more infrastructure to monetize. Selling unused dark fiber capacity to Google and other AI-focused tech companies could help recoup some of the acquisition costs, according to the source article.
Central Offices Become Small Data Centers
Verizon is not limiting AI Connect to long-distance data center links. The company is also in the early stages of retrofitting some of its central offices into small data centers.
Those central offices are locations that house physical equipment for Internet and mobile networks. Verizon has already been removing copper wiring from these sites and upgrading them with fiber-optic technology.
That work began before the company looked at converting some offices into AI infrastructure for edge computing. Verizon has been retiring its legacy copper network in favor of faster and more power-efficient fiber-optic cables for broadband services.
The new use case is “inference edge computing.” In plain language, Verizon wants some AI workloads to run closer to where services are used, rather than only inside large centralized data centers.
Schulman tied this to applications that need ultra-low latency, including robotics, remote surgery, autonomous driving, autonomous robotaxi fleets, and robotics applications. He said there is demand to “move that out into the edge, as opposed to these big, massive data centers.”
Early Demand and Long-Term Revenue Expectations
Verizon says it has already tested the central-office concept on a small scale. Schulman said, “We did a small trial on that and sold out capability in 24 hours.” He added, “So we’re seeing a large demand for that as well.”
The company is not presenting AI Connect as an immediate replacement for its core businesses. Schulman said the deals are expected to “begin to impact our revenues and margins beginning next year and grow substantially over the next five to 10 years.”
That would sit alongside Verizon’s core mobility and broadband services, which saw record second-quarter results this year. The AI infrastructure effort is therefore being framed as an additional growth path, built on assets Verizon already owns or is already upgrading.
At the same time, the company has been cutting jobs. Verizon laid off 13,000 workers at the end of 2025, leaving a workforce of just under 90,000 employees at the end of last year. More recently, it transferred 274 retail stores to franchise owners along with 2,500 employees and cut another 500 positions within its own corporate structure.
What the Strategy Signals
Verizon’s AI Connect strategy connects several strands of the company’s business: dark fiber, expanded fiber coverage, copper retirement, broadband upgrades, and central-office real estate. The common thread is that AI demand may create new revenue from network infrastructure that was built for communications but can also support compute-heavy services.
The Google deal gives Verizon a concrete proof point. The central-office trials suggest a second path, focused on low-latency inference at the edge.
For now, Verizon has disclosed one major dark fiber customer and early demand for remote data center capacity. The company’s next test is whether the additional AI-related deals it expects by the end of the year turn AI Connect into a larger revenue contributor over the next several years.