AI Infrastructure Buildout Faces Major Delays Into 2027

AI Industry Expansion Timeline Extends as Infrastructure Lags

The artificial intelligence industry should prepare for substantial infrastructure delays extending through 2026 and 2027, according to Wes Cummins, CEO of Applied Digital, a company focused on AI infrastructure and data center operations.

Cummins shared his outlook on the AI buildout during an appearance on CNBC's "Squawk on the Street," where he discussed the challenges facing data center development and the broader implications for companies racing to deploy AI capabilities at scale.

Why it matters

These projected delays signal a potential bottleneck in AI adoption that could affect everything from enterprise deployment timelines to competitive positioning among tech companies. Organizations planning AI initiatives may need to adjust their roadmaps and investment schedules, while the gap between AI demand and available infrastructure could create pricing pressures and strategic advantages for companies that secured capacity early.

Infrastructure Constraints Slow AI Momentum

The warning from Applied Digital's leadership comes as the AI industry grapples with unprecedented demand for computational resources. Data centers—the physical backbone required to train and run large AI models—have become a critical constraint as companies compete for limited capacity.

Applied Digital operates in the intersection of these challenges, providing high-performance computing infrastructure specifically designed for AI workloads. Cummins' perspective offers insight from a company directly involved in building the infrastructure that AI development depends on.

The delays Cummins anticipates likely stem from multiple factors common in large-scale infrastructure projects: permitting processes, power grid connections, equipment procurement, and construction timelines. Data centers require massive amounts of electricity and cooling capacity, often necessitating utility upgrades that can take years to complete.

Implications for the AI Ecosystem

For AI companies and enterprises planning deployments, these delays could mean extended wait times for accessing the computational power needed to train models or serve applications to users. The constraint may particularly impact smaller companies and startups that lack existing infrastructure relationships or the capital to build their own facilities.

Larger technology companies with established data center operations or early commitments to capacity may find themselves with a competitive advantage during this constrained period. The timeline Cummins describes—stretching through 2027—suggests the infrastructure gap won't resolve quickly even as new facilities come online.

The details were first reported by CNBC during Cummins' interview on the network's morning programming.