Skip to main content

Connecting Our AI Future

Connecting Our AI Future

As the AI economy accelerates, infrastructure leaders like Lumen are stepping into a pivotal role—one that blends deep expertise in networking with bold strategic moves to power its enterprise, multi-cloud AI-first world strategy. During a live broadcast of Fiber for Breakfast during Fiber Connect 2025, James Feger, representing Lumen, shared key insights on how the company is doubling down on its AI-first strategy and reimagining the fiber and cloud landscape for an increasingly data-driven world.

A Strategic Pivot Toward Enterprise and AI

Lumen’s recent decision to sell off its mass-market residential fiber business to AT&T marks more than just a portfolio reshuffle—it signifies a major strategic refocus. By exiting the consumer broadband market and honing in on AI infrastructure and enterprise services, Lumen is channeling resources into its multi-cloud AI-first world strategy and AI-focused infrastructure strategy. Backed by over $5 billion from the divestiture, this realignment allows the company to intensify investments in high-performance, scalable networks optimized for AI workloads.

“Our strategy is all about focus,” said Feger. “We’re investing heavily in infrastructure to support the AI economy—not just talking about it, but putting money where our mouth is.”

Why AI Infrastructure Now?

According to Feger, Lumen’s confidence in its AI-first approach is based on tangible momentum across industries. Whether working with hyperscalers, banks, or manufacturers, the consistent demand for low-latency, high-throughput connectivity is undeniable.

“We’ve reached a point where real-world use cases are proliferating,” Feger said. “Every time we build something and see it get consumed immediately, that’s validation. It’s no longer hype—this is the new reality.”

This trend is visible in the scale of Lumen’s networking contracts, totaling nearly $9 billion in AI and cloud-related partnerships. These efforts reflect not only growing demand, but also the company’s belief that AI will reshape network architectures as we know them.

The Role of Hyperscalers and Edge

AI’s explosive growth is largely fueled by hyperscalers—big tech players like Microsoft, Meta, Amazon, and Google—who are building massive data centers that act as the foundation for AI model training and deployment. Lumen plays a key role in enabling this ecosystem by ensuring seamless connectivity between these high-powered compute environments.

“Think of these data centers as foundational hubs,” Feger explained. “Our job is to build the highways between them and out to the edge—whether that’s another data center, a corporate campus, or an industrial site. That’s where fiber comes in.”

Fiber: The Lifeblood of AI

Feger emphasized that fiber is absolutely critical to AI’s future, providing the scale, latency, and bandwidth needed to support today’s compute-heavy applications. Lumen’s network has seen exponential increases in capacity demands, with hundreds of high-speed circuits being deployed between data centers.

“We’re pushing terabits of traffic across fiber routes,” Feger said. “From an engineering standpoint, it’s staggering—and it’s just the beginning.”

In addition to investing in traditional fiber builds, Lumen is exploring next-gen technologies like hollow-core fiber, which could further reduce latency and increase efficiency. Innovations in fiber strand density and streamlined installation (such as blowing fiber through existing conduit) are also helping to meet demand without massive trenching operations.

Strategic Edge Deployment & Power Considerations

As AI workloads become more distributed, building closer to sources of excess power is becoming a strategic necessity. Countries like Singapore and regions like Kuala Lumpur are already experimenting with placing AI compute closer to energy sources, and fiber is the key enabler.

“We need to plan builds not just to connect cities,” Feger said, “but to reach remote areas where power is abundant. That’s where the next data center may land.”

This shift toward edge computing requires a rethink of traditional fiber strategies. Instead of long-haul builds that connect distant cities, operators now have to focus on how many viable pickup points they can enable along the way—balancing cost, resilience, and expansion potential.

Data as the New Oil—and How to Move It

AI’s potential hinges on data—and moving massive volumes of it quickly and securely is a new challenge. From petabytes of stored data to AI clusters across the globe, Lumen is seeing a sharp increase in large-scale data transfers that strain existing network infrastructure.

“Moving AI training data from storage to processing is one of the biggest bandwidth drivers today,” Feger noted. “Sometimes we’re talking about transferring terabytes daily over dedicated, high-capacity links. Fiber is the only viable solution.”

Preparing for Quantum and Secure Networks

Security is also top of mind, particularly with the looming potential of quantum computing. Lumen is already implementing quantum-resistant encryption in key parts of its network and working with government and enterprise clients on secure networking solutions.

“Quantum networking is no longer theoretical,” said Feger. “We’re actively preparing for that future.”

The Road Ahead

Lumen’s vision is clear: build an AI-native, multi-cloud network fabric that connects every source to every destination with unmatched reliability and speed. That future depends on forward-thinking design, deep customer collaboration, and a relentless commitment to innovation—especially in fiber.

“AI doesn’t happen without fiber,” Feger concluded. “It’s not just important—it’s absolutely foundational.”

Click here to listen to the full episode or find previous episodes of Fiber for Breakfast.

More resources:
Keep updated on Fiber Connect 2026