Verizon sees AI data center transformation ahead as rivals eye inference

Verizon has inked a deal worth more than $1 billion with Google to connect data centers with its dark fiber assets and is retooling its central offices to be data centers for AI inference.

The Verizon moves highlight how telecommunications companies could become key AI infrastructure players because they already have the footprint of facilities. While AI infrastructure has been focused on building new facilities, there's an argument to be made that much of the needed compute could be in place already.

• What if our AI infrastructure is already built?

Dan Schulman, CEO of Verizon, said on the company's second quarter earnings call that the plan is to become more than just a wireless carrier. He said the Google deal is the first of a series.

"We've been building carrier-grade fiber routes for decades. We know how to get it done. We understand permitting. We do our own construction. And we have a solid balance sheet and a solid service reputation that the AI ecosystem needs and counts on," said Schulman.

Schulman added that its central offices are being retooled for AI inference. Verizon has thousands of central offices where copper networking is being retrofitted for remote data center networking that's already power-ready and permitted.

He said:

"We are in the early stages of retrofitting many of our central offices into data centers for inference edge computing with multiple conversations underway with partners who are eager to utilize these power-ready and permitted locations. We are moving quickly to expand our TAM in the rapidly growing AI infrastructure market. The agreements we have signed are the leading edge of a strategy that will become a meaningful incremental leg of growth for Verizon."

Schulman said the AI infrastructure market is evolving from optimizing compute to networking and connectivity from the data center to the edge.

"While data center connectivity demand is exploding, there's an equal amount of desire and demand to power inference models and applications that need ultra-low latency into the edge as opposed to these big, massive data centers," said Schulman.

• Verizon aims for autonomous network: Takeaways on architecture, AI agent returns, managing costs
• How Verizon manages AI agent sprawl
• Verizon launches AI Connect, courts AI inferencing workloads
• Verizon bets on AI experiences, Google Cloud to bring down churn

Here's a look at how Verizon's rivals are targeting AI workloads.

AT&T

AT&T has focused on its base of converged customers that get home internet as well as wireless service. The telecom is focused on AT&T Fiber services as well as its fixed wireless offering.

The company's enterprise unit is also focusing on the combination of fiber and 5G.

CEO John Stankey said the strategy revolves around building a network for AI traffic. He said:

"The proliferation of agentic and autonomous AI workloads will require networks to sense, decide and act in near real time. Emerging use cases, including drones, autonomous driving, robotics and AR glasses will all require ubiquitous high-performing uplink optimized connectivity. Today, industry research shows AI agents generate up to 450% more total traffic per task than a human performing the same work. Agentic adoption is projected to drive approximately 9x growth in enterprise traffic and approximately 7x growth in consumer traffic by 2035. Distribution of AI inference to the edge necessitates low latency and high-bandwidth connectivity to access endpoints. This is why we believe fiber-enabled network convergence at the edge will create a true competitive advantage."

• Constellation ShortList™ Infinite Connectivity Vendors

T-Mobile

T-Mobile US is primarily focused on wireless and CTO John Saw said AI traffic today is focused on data center workloads, but expects that will change.

Saw said AI will be about edge inference and physical AI for T-Mobile. He said:

"With the growth of physical AI, I think we are more than prepared and we have sufficient runway to continue to invest in our capacity, both in terms of existing spectrum, and we haven't even looked at the impact of getting new spectrum. We rolled out 5G Advanced recently and one of the reasons why we do that is that we want to improve the uplink performance, especially with the surge in -- potential surge in AI mobile traffic soon."

Srinivasan Gopalan, T-Mobile CEO, said "we believe our network will become the connective tissue for physical AI with inferencing at the edge."

He added that 6G will be about edge AI and physical AI as well as a big boost for wireless networks.