Predictions from Zayo, Digital Infrastructure, and Mission Critical Group about what to expect from AI infrastructure in 2026
As AI workloads scale and infrastructure cycles compress, the pressure on networks, power systems, and physical deployment models is intensifying. By 2026, industry leaders expect competitive advantage to hinge less on raw compute and more on how intelligently infrastructure can be designed, connected, secured, and evolved.
Executives from Zayo, Digital Infrastructure, and Mission Critical Group outline how AI-driven security, network-centric architectures, long-haul fiber consolidation, power system adaptability, and regulatory and labor constraints will shape the next phase of AI infrastructure. RCR has compiled their predictions below.
Shawn Edwards, chief security officer, Zayo:
By 2026, the resilience of a network will be defined by how intelligently it can defend itself
“Traditional, static network architectures simply can’t keep pace with AI-accelerated threats that mutate, move laterally, and probe infrastructure faster than human operators can respond. The advantage shifts to networks that can sense and interpret abnormal patterns in real time, automatically adjust routing, prioritize clean paths, and quarantine hostile traffic before it impacts services. As adversaries lean into autonomous AI to discover weaknesses at scale, the only viable defense will be networks that use AI to continuously learn, adapt, and harden themselves. The modern network won’t just move packets; it will detect, predict, and respond at machine speed.”
Jason ‘JJ’ Jorgensen, senior vice president of network implementation, Zayo
Despite insatiable bandwidth demand, the long-haul ecosystem will start to narrow — for the better.
“The surge of new capital into long-haul fiber will spark innovation and much-needed expansion, but not every entrant will be equipped for the operational realities of large-scale builds. The gauntlet of permits, jurisdictions, vendor chains, and environmental factors can derail a project that already costs upwards of tens of millions of dollars, leaving some developers in the red and delaying critical progress. Over the next two years, we’ll see the market self-correct as experience and execution rise to the top. At the same time, the sheer pace and scope of construction will force deeper aggregation across the ecosystem to accelerate delivery at scale. This natural shaking-out of the market will create a more disciplined ecosystem capable of building at the scale, speed, and quality AI requires.”
James Tomko, senior vice president, Digital Infrastructure
The AI infrastructure stack flips
“By 2026, the network will define AI performance. AI training, inference, and data movement will stretch across regions and regulatory boundaries, and the real limiter won’t be GPUs but interconnects across the entire AI ecosystem. As distributed AI fabrics emerge, success will depend on how intelligently data moves between compute nodes, not just how fast it’s processed inside them. As such, the network will become the control plane of AI.”
By 2026, the competitive edge in AI won’t come from compute density alone, but from network design
“As AI workloads scale across distributed data centers, the ability to move, synchronize, and manage data efficiently will matter as much as raw compute. Metro-scale and long-haul fiber will define the winners of distributed AI — those who can interconnect and orchestrate data across regions, clouds, and edges. The next wave of AI leadership won’t be won in the data center alone, but across the networks that connect them.”
Jeff Drees, CEO, Mission Critical Group
Electrification and GPUs will reshape the design cycle, forcing continuous evolution
“Organizations building or integrating power systems within industries undergoing electrification — such as data centers, automotive, manufacturing, and healthcare — must realize that today’s designs will be obsolete within two to three years. Rapid changes in GPU and CPU technology, as well as evolving cooling approaches, are reshaping power and electrical architectures at unprecedented speed. The old model of over-engineering with a buffer is disappearing; new sites must be built for immediate high-density performance with minimal slack.
As a result, engineering organizations will embrace modular blocks, upgradeable components, factory testing, and use scenario-based planning to prepare for future density increases. Designs will shift from static, single-lifecycle projects to flexible systems built for repeat refreshes across a 20- to 30-year building lifespan. Iteration will become a daily reality as new compute requirements ripple through cable sizing, voltage strategy, and power delivery pathways. In 2026, adaptive engineering practices will be the only viable response to a world where infrastructure must evolve as fast as the technology it supports.”
Matt Coffel, chief commercial and innovation officer, Mission Critical Group
Regulatory and permitting friction will slow the race to build AI infrastructure
“Regulation and permitting will remain a defining constraint on data center expansion in the United States throughout the coming year and beyond. As operators and technology companies race to scale capacity amid accelerating AI demand, they will continue to confront challenges with securing approvals from local governments and state agencies, not only for building permits but also for environmental impacts, emissions, and noise.
And, because regulations and permitting requirements vary widely by jurisdiction — where the process can differ dramatically within major cities versus just outside city limits or in more rural areas — businesses must learn to navigate a maze of rules that introduce complexity, uncertainty, and potential delays to their projects. As a result, data center operators will increasingly seek partners who understand local regulatory landscapes and can deliver solutions tailored to the specific requirements of the communities where they plan to build.”
Skilled labor shortages will become a critical risk in the data center supply chain
“With the growing development of data centers — and the critical power infrastructure required to support them — there is an ever-increasing demand for skilled labor such as electricians, welders, HVAC technicians, and even plumbers. However, reports indicate that a gap between available talent and open jobs will develop, increasing risks for data center operators. This demand for skilled workers will push organizations, governments, and educational institutions to accelerate efforts to train the next generation for these roles. More specifically, we will see deeper collaboration between employers and nearby educational institutions — particularly those located near power-infrastructure manufacturing hubs, data center build sites, and similar industrial corridors — to create programs tailored to the technical requirements of these jobs and establish clear, path-to-career pipelines for students upon graduation. We’ll also see data center operators and power-supply partners lean more heavily on complete, pre-tested power modules that simplify installation and help offset the growing strain on specialized labor.”
