While cloud providers continue expanding compute capacity, reliable connectivity has become just as essential as the graphics processors powering large language models. That shift is creating new opportunities for carriers with extensive transport assets and decades of infrastructure investment behind them. Verizon’s latest commercial agreement with Google illustrates how telecommunications operators are increasingly becoming strategic infrastructure partners rather than traditional connectivity providers. The transaction also highlights how hyperscale AI deployment now depends on networks capable of moving enormous volumes of data between geographically distributed compute environments. Instead of competing solely on bandwidth, carriers are beginning to compete on resilience, latency, scalability and long-term infrastructure readiness.

Google Partnership Signals AI Infrastructure Priorities

Verizon disclosed during its second-quarter earnings discussion that it signed a $1 billion dark fiber agreement with Google, representing one of the company’s most significant publicly disclosed AI-focused network agreements. Rather than representing a conventional telecommunications contract, the agreement reflects growing demand for dedicated optical networks linking hyperscale AI facilities across multiple regions. Dark fiber has become increasingly valuable because operators can dedicate entire strands to high-capacity workloads without sharing network resources with other customers. The agreement also demonstrates how hyperscalers now prioritise network ownership characteristics alongside compute availability when planning AI expansion. Meanwhile, Verizon positioned the agreement as evidence that its AI infrastructure strategy is beginning to translate into large-scale commercial execution. The development reinforces a broader industry trend in which transport infrastructure has become a strategic layer within artificial intelligence ecosystems instead of remaining a background utility.

Connectivity Network

Verizon executives linked Google’s agreement to the company’s AI Connect initiative, a platform launched under former Chief Executive Hans Vestberg during 2025 to address emerging infrastructure requirements surrounding artificial intelligence deployments. The initiative combines dark fiber, programmable networking, cloud connectivity, GPU-as-a-service integration and cybersecurity capabilities into a unified enterprise offering. Instead of offering individual network services, Verizon positions AI Connect as an integrated platform that combines communications infrastructure and AI-focused connectivity capabilities for hyperscale customers. That approach aligns with growing enterprise demand for providers capable of supporting increasingly complex AI deployment models through integrated connectivity and infrastructure services. Network availability alone is no longer sufficient because AI infrastructure requires coordinated transport, security and cloud connectivity across multiple locations. Verizon believes that combining those capabilities strengthens its competitive position as hyperscale infrastructure investment accelerates across North America.

Dan Schulman described the commercial significance of the Google agreement during the earnings discussion while outlining Verizon’s broader growth strategy. “Quite consequential because it foreshadows where our revenue growth profile is going from here.” His comments suggest the company views artificial intelligence infrastructure as a long-term growth engine rather than an isolated enterprise opportunity. Schulman also indicated that Verizon’s customer pipeline extends beyond a single hyperscale engagement, reflecting broader market demand for dedicated AI connectivity. “We have other deals that we expect to announce by year-end that, taken together, are expected to be worth multiple billions of dollars in revenue over the next several years,” he said. “These are long-duration, high-quality contracted revenue streams from some of the most demanding infrastructure customers in the world.” Those remarks indicate that Verizon expects AI connectivity to become an increasingly important contributor to its portfolio of long-term contracted enterprise revenue.

Fiber Networks Become Strategic AI Assets

The Google agreement arrives as telecommunications providers increasingly compete to support hyperscale AI deployments requiring extensive optical transport infrastructure. Building AI data centres represents only one element of the infrastructure equation because those facilities must also exchange massive datasets across metropolitan and long-haul networks with extremely low latency. As a result, fibre capacity has become a strategic differentiator for carriers serving cloud providers expanding regional AI clusters. Rather than waiting for demand to emerge, several infrastructure operators have already secured long-term supply agreements and expanded backbone routes supporting AI growth corridors. Verizon’s announcement therefore reflects wider structural changes rather than an isolated commercial success. The telecommunications sector is steadily evolving into a critical participant within the artificial intelligence infrastructure supply chain.

Industry demand indicators reinforce the strategic importance of optical transport within artificial intelligence ecosystems. Zayo’s second annual Bandwidth Report noted that demand for long-haul dark fibre doubled between 2024 and 2025 as its largest customers expanded interconnections between AI data centres supporting substantially larger computational workloads. Those trends illustrate how AI deployment extends far beyond server procurement because every new compute cluster requires resilient network pathways linking distributed processing resources. Furthermore, dedicated optical infrastructure enables hyperscale operators to maintain predictable performance while accommodating rapidly growing data movement requirements. Such demand dynamics continue encouraging telecommunications providers to invest in additional transport capacity despite broader economic uncertainty. Fibre networks are increasingly emerging as foundational assets underpinning the next generation of artificial intelligence infrastructure.

Verizon Expands Its AI Connect Strategy

Verizon enters that competitive environment with one of North America’s largest owned transport networks spanning more than 500,000 route miles across long-haul and metropolitan infrastructure. The company assembled that footprint through decades of organic expansion alongside acquisitions including MCI and Frontier, providing substantial geographic reach across key enterprise markets. Verizon’s extensive network footprint provides broad geographic coverage that can support hyperscale operators requiring connectivity across multiple regional markets. Existing fibre assets provide established transport infrastructure that can support customers expanding AI deployments across multiple metropolitan regions. Verizon is expanding its network infrastructure as it prepares to support continued enterprise demand for AI connectivity over the coming years. The strategy reflects increasing customer preference for integrated infrastructure partners capable of delivering nationwide connectivity at scale.

Schulman argued that Verizon’s infrastructure portfolio uniquely positions the company to benefit from this investment cycle. “The build-out of AI infrastructure across the United States is one of the largest capital cycles of our lifetime and Verizon is uniquely positioned to participate in it,” Schulman said. “We own one of the most extensive long-haul and metro fiber footprints in North America. We have spent decades building the kind of carrier-grade, low latency, highly resilient transport network that hyperscalers need to connect compute to compute, model to model, and region to region.” His comments emphasise infrastructure quality rather than simple network scale. Verizon appears determined to position itself as a foundational infrastructure provider supporting increasingly sophisticated AI ecosystems. That narrative aligns with growing enterprise demand for resilient connectivity capable of supporting continuously expanding computational workloads.

Legacy Infrastructure Creates New Data Centre Opportunities

Verizon’s artificial intelligence strategy extends beyond optical transport into infrastructure modernisation across its traditional telecommunications estate. Like AT&T, the company continues retiring legacy copper networks that historically supported its longstanding public switched telephone services for residential and enterprise customers. Replacing copper with fibre has reduced operational complexity while enabling Verizon to consolidate equipment previously occupying large multi-storey central office facilities. According to the company, infrastructure modernisation has reduced the energy requirements associated with real estate and cooling by approximately sixty to eighty percent. Those operational improvements also create opportunities to repurpose underutilised facilities for new infrastructure functions as Verizon continues modernising its network. Consequently, legacy telecommunications assets are gradually being repositioned to serve emerging digital infrastructure markets.

Although Verizon sold most of its dedicated data centre portfolio to Equinix in 2017, management believes its central office footprint now represents a valuable strategic resource supporting artificial intelligence deployment. Existing facilities already possess power availability, permitting and established network connectivity, reducing several barriers typically associated with new data centre development. Instead of constructing entirely new campuses, Verizon can potentially repurpose infrastructure originally designed for traditional telecommunications operations into distributed edge computing environments. That approach aligns with broader industry interest in bringing AI processing closer to users while leveraging existing infrastructure wherever practical. “We have thousands of central offices, many of which were taking copper out of, and we are retrofitting them to be remote data centers that are power-ready, permitted, fully redundant infrastructure,” he said. If executed successfully, the strategy could transform legacy telecommunications assets into an important component of the United States’ expanding AI infrastructure landscape.