Verizon Google AI deal tops $1 billion as data center demand rises
Verizon says Google will use its dark fiber to link data centers, with more AI infrastructure contracts expected by year-end.
By Rafael Ortiz · Fintech Correspondent
· 3 min read
Verizon signed a Verizon Google AI deal worth more than $1 billion to support data center connectivity, CEO Dan Schulman said Friday during the company’s second-quarter earnings call. The agreement comes as large technology groups increase capital spending on artificial intelligence infrastructure and need high-capacity networks to connect computing sites.
Under the agreement, Google will use Verizon’s dark fiber to connect its data centers, Schulman said. Dark fiber refers to fiber-optic cable that has been installed but is not yet carrying traffic, allowing a customer to light and operate capacity for its own high-volume network needs.
Schulman said Verizon expects to announce several similar agreements before the end of the year. Those contracts would generate multiple billions of dollars in revenue over the next several years, according to his comments on the call.
What is the Verizon Google AI deal?
The deal is a contracted infrastructure arrangement in which Verizon provides fiber capacity that Google can use to link data centers. For AI operators, these connections help move data between computing clusters, models and regional facilities with low latency and high reliability.
Schulman described the contracts as long-term revenue streams with demanding infrastructure customers. He said Verizon views the build-out of AI infrastructure across the United States as one of the largest capital investment cycles of the current era.
Verizon’s case rests on its existing long-haul and metro fiber network, according to Schulman. Long-haul fiber connects distant regions, while metro fiber links sites within major urban areas. Together, those assets can serve cloud providers and other hyperscale customers that need dedicated transport capacity between data centers.
The company is also converting many of its central offices into data centers for inference edge computing, Schulman said. Inference is the stage in which an AI model is used to generate outputs after training, and edge locations can place computing power closer to users or enterprise sites. Schulman said Verizon is already in discussions with several partners interested in those power-ready and permitted locations.
How Verizon is positioning its AI infrastructure business
Verizon introduced Verizon AI Connect in a January 2025 announcement, describing it as a strategy and product set for hyperscalers, cloud providers and global enterprises that need to handle AI-intensive workloads. At the time, the company named Google Cloud and Meta as early adopters.
In its Friday earnings release, Schulman said Verizon’s core connectivity business was gaining momentum and that AI infrastructure revenue was changing the company’s growth profile. He also said the agreements already signed were the start of a strategy that Verizon expects to become a meaningful additional source of growth.
The Google agreement also sits within a broader spending cycle among major cloud and AI companies. PYMNTS reported Wednesday that Alphabet, Google’s parent company, raised its 2026 capital spending forecast during its second-quarter earnings call. Alphabet lifted the range from a prior forecast of $180 billion to $190 billion to a new range of $195 billion to $205 billion, according to that report.
For Verizon, the immediate market signal is that AI demand is extending beyond chips and servers into fiber transport, power-ready sites and data center connectivity. Schulman said the company expects more contracted AI infrastructure revenue as customers build larger distributed computing networks.
This story draws on original reporting from PYMNTS.
