Verizon secures $1 billion Google dark fiber deal for AI infrastructure
Verizon has entered into a $1 billion agreement to provide Google with dedicated dark fiber infrastructure to support hyperscale AI workloads and data center interconnects. The deal reflects a strategic move by telecom providers to monetize long-haul network assets by enabling low-latency connectivity for AI clusters.
Key Takeaways
- Google gains operational control over the optical layer for custom traffic engineering and deterministic latency.
- Verizon CEO Dan Schulman confirmed a multi-billion-dollar pipeline of similar hyperscale contracts expected by year-end.
- The contract utilizes Verizon's long-haul and metro network as a backbone for distributed AI clusters.
- Verizon has begun converting legacy central offices into edge data centers specifically for AI inference.
Why It Matters
The shift from managed services to dark fiber leasing represents a strategic pivot for telecom providers seeking to monetize legacy transport assets in the high-growth AI infrastructure market. For companies like Google, dedicated fiber is a non-discretionary requirement for scaling compute-to-compute data transfer across geographically distributed campuses. This move highlights a broader trend where telecom real estate, including power-ready central offices, is being revalued as critical edge compute nodes for low-latency AI inference. As fiber assets and grid-ready sites become bottlenecks for hyperscalers, expect a surge in long-duration infrastructure-level partnerships. Watch for Verizon to announce an additional major hyperscale contract before the close of 2026 to validate this multi-billion-dollar revenue trajectory.
Additional Context
The Google agreement follows Verizon's January 2025 long-haul fiber deal with AWS, part of the company’s ‘AI Connect’ initiative. This strategy focuses on retrofitting legacy assets—including the One Fiber infrastructure and 5G network slicing—to manage resource-intensive generative AI workloads. Recent industry data underscores this shift; per the International Energy Agency (IEA) in early 2026, global data center electricity consumption is projected to double by 2030, making pre-permitted, power-ready telecom facilities like Verizon’s central offices highly valuable to cloud providers facing long grid interconnection queues. According to McKinsey research cited by Verizon, roughly 60-70% of AI workloads will transition to real-time inference by 2030, necessitating the ultra-low latency provided by these edge computing AI demand conversions.
Competitively, the dark fiber market is reaching a permanent reset as hyperscalers increase capital intensity. Per Goldman Sachs (May 2026), the top five hyperscalers are projected to spend over $600 billion on infrastructure in 2026, with 75% targeted specifically toward AI. Simultaneously, rivals like Zayo reported that long-haul dark fiber demand doubled between 2024 and 2025 as customers expanded connectivity to support larger AI clusters. Verizon's scale was recently boosted by its January 2026 acquisition of Frontier Communications, which brought its total fiber footprint to roughly 30 million passings. This integration provided the necessary density to compete for the 'unmanaged' transport deals favored by Google and Amazon, which require high-capacity paths between increasingly power-constrained metros like Northern Virginia and Dallas-Fort Worth, a trend further complicated by synchronized tiered egress pricing models across major cloud providers.
Read full article at submarinenetworks.com
Enjoy our coverage?
Add StreamingMeme as a preferred source on Google to see more of our streaming news at the top of your Search results.
Add as preferred source