Zayo has released its second annual Bandwidth Report, examining bandwidth consumption trends among nearly 6,000 customers throughout 2025. The report identifies increased investment in high-capacity networks as organizations adapt their infrastructure strategies to support artificial intelligence (AI) workloads.

According to the report, AI is influencing network planning decisions, including capacity requirements, deployment locations, and network architectures needed to connect data centers, GPU clusters, cloud regions, and end users across a more distributed computing environment.

The findings indicate that organizations preparing their networks for AI-driven growth are increasing investment in connectivity infrastructure, while those relying on networks designed primarily for traditional workloads may face limitations as demand for higher-capacity connectivity continues to grow.

“AI is exposing a hard truth: networks built for yesterday’s traffic patterns will not support tomorrow’s workloads,” said Bill Long, Chief Product & Strategy Officer at Zayo. “The organizations set up to capture AI’s value are investing now in the bandwidth, fiber, and architecture required to scale. They understand that the next decade of competitive advantage will depend on network infrastructure that is built for AI.”

Key Findings from the Report

AI adoption continues to influence network investment strategies: Bandwidth investments in 2025 continued at the record levels observed in 2024 following broader AI adoption. Demand for wavelengths and metro dark fiber remained strong as organizations expanded infrastructure to support higher-capacity applications.

Metro dark fiber demand increases as AI processing moves closer to users: The report found significant increases in metro dark fiber demand in markets experiencing AI-related growth. In some major AI-focused metropolitan areas, demand increased by as much as 20 times as organizations shifted from AI model training toward inference workloads that require processing closer to end users.

Long-haul dark fiber demand grows alongside larger distributed workloads: Demand for long-haul dark fiber doubled between 2024 and 2025 among Zayo’s largest customers as organizations expanded connectivity between AI data centers to support increasingly complex AI workloads.

Enterprise networks face pressure from AI and digital transformation initiatives: Wavelength capacity purchases increased among several customer segments, including a 2.6x increase among data centers, a 3.2x increase among media and internet providers, and a 2.3x increase among financial institutions.

Hyperscalers and carriers remain major buyers of high-capacity fiber infrastructure: The report found that hyperscalers and carriers accounted for 95% of long-haul fiber purchases. AI infrastructure providers and neocloud companies are also emerging as significant buyers of high-capacity network solutions as demand grows for connectivity among data centers, GPU clusters, and cloud regions.

Network Infrastructure Becomes Increasingly Important to AI Deployment

As organizations expand AI adoption, network capacity is becoming a key consideration alongside computing resources. The relationship between AI and infrastructure reflects a broader shift toward distributed digital environments, where connectivity plays an important role in moving data between applications, computing resources, and users.

Organizations using infrastructure designed for traditional workloads may encounter challenges related to AI performance, deployment timelines, and operational efficiency. Companies investing in higher-capacity networks are positioning their infrastructure to support evolving AI workloads.

To address anticipated demand, Zayo is expanding its fiber infrastructure, including more than 15,000 route miles of long-haul fiber development across North America. The company has also expanded its metro network capabilities through the acquisition of Crown Castle’s Fiber Solutions business, adding approximately 90,000 route miles and 40,000 on-net enterprise locations.

The investments reflect broader industry efforts to expand digital infrastructure capacity as AI adoption drives new requirements for connectivity, data movement, and network scalability.

Source: Zayo

Stay Connected with ISE Magazine 

Subscribe to our newsletters and magazine for the latest telecom insights, explore the current issue for in-depth features and strategies, and register for upcoming webinars to learn directly from industry leaders.