Fiber for Breakfast Week 23: State of I&O Strategy in AI Era 2H 2026

Fiber for Breakfast Week 23: State of I&O Strategy in AI Era 2H 2026 

For the past two years, AI has dominated technology conversations. But for many organizations, the focus has been on experimentation—testing tools, exploring use cases, and trying to understand where AI fits into the business. This week’s Fiber for Breakfast suggested that phase may be ending.  

Joining Gary was Ron Westfall, Vice President and Practice Leader for Infrastructure and Networking at HyperFRAME Research, who shared findings from the firm’s latest global Infrastructure & Operations (I&O) study. Based on input from more than 500 decision-makers worldwide, the research shows organizations shifting from AI exploration to AI deployment.  

That shift is creating a new set of challenges. The conversation is no longer centered on what AI can do, but whether the underlying infrastructure is ready to support it.  

One of the strongest themes throughout the discussion was the changing nature of network traffic. Broadband networks have traditionally been designed around north-south traffic flows, moving information between users and centralized destinations. AI is beginning to alter that model.  

“We’re seeing a shift in the traffic that must be spotlighted,” Westfall said. “East-west flows within data centers are definitely expanding dramatically.” 

AI workloads depend on constant movement of data between servers, clusters, and facilities. Training models, synchronizing datasets, and supporting inference all depend on high-capacity connectivity. As a result, demand is growing not only at the edge of the network but also deep inside data centers and across regional and national backbone infrastructure.  

For the fiber industry, that represents an important evolution. The industry’s recent investment cycle has focused heavily on connecting homes and businesses. AI introduces another layer of demand, one that places increasing importance on middle-mile networks, backbone capacity, and data center interconnection. 

The research also highlighted a reality many organizations are already experiencing: adopting AI is proving more complicated, and more expensive, than expected. 

Security and compliance ranked as the top concern among respondents, while cost management and infrastructure readiness followed closely behind. Many organizations reported that AI investments have exceeded initial projections, creating pressure to balance innovation goals with operational and financial realities.  

Yet despite those challenges, few organizations appear to be slowing down. More than half of respondents expect to increase investments in virtualization and cloud infrastructure, highlight how quicky AI is moving from an emerging technology initiative to a core infrastructure priority.  

Taken together, the findings point to a market that is moving beyond AI experimentation and toward real-world deployment. The focus is shifting from pilots and proofs of concept to the infrastructure, processes, and network architectures needed to support AI at scale.  

For fiber providers, that’s a significant development. As AI workloads become more distributed and data-intensive, the network itself becomes increasingly important to overall performance.  

The organizations making infrastructure decisions today aren’t simply preparing for higher bandwidth demand. They’re building the foundation that will determine how effectively AI can be deployed and scaled in the years ahead.  

Click here to watch the full interview.