AI ambition is crashing into a decade of deferred IT maintenance, warns Red Hat CEO

Enterprise AI infrastructure modernization has reached a critical crossroads as organizations grapple with decades of technical debt while facing intense pressure to deploy AI — and returning to IT fundamentals is now the only viable path forward.

Nowhere is that collision more visible than at the intersection of platform engineering and AI-driven infrastructure, where organizations are discovering that the race to deploy AI is won or lost long before the first model hits production. Enterprise AI infrastructure modernization is now nonnegotiable, as boards are demanding returns on AI investment while IT teams get buried under complexity they never fully resolved, according to Matt Hicks (pictured), president and chief executive officer of Red Hat Inc.

“From the board level down … you have a feeling of, ‘We have to be able to harness what this technology can do,'” Hicks said “They apply that top-down pressure, and then if you are in the IT team, for the last 10 years, we have been trying to get to cloud or ending up in multiple spots, dealing with tons of technologies, but not always maintaining technical debt. That is all crashing down on them at the same time.”

Hicks spoke with theCUBE’s Rob Strechay and Rebecca Knight at Red Hat Summit 2026, during an exclusive broadcast on theCUBE, SiliconANGLE Media’s livestreaming studio. They discussed enterprise AI infrastructure modernization, the organizational pressures reshaping IT teams and how the future of engineering work is being redefined by AI agents. (* Disclosure below.)

Enterprise AI infrastructure modernization demands a return to basics

The tension between AI ambition and infrastructure reality is showing up most acutely in how enterprises handle the systems they already operate. As Red Hat expands its agentic AI strategy with new inference, automation and governance capabilities, the company is working to steer customers toward closing the maintenance gap before chasing new capabilities, Hicks explained.

“We’re in a back-to-basics mode where you have to learn how to patch your systems. One year ago, you might have argued the risk was okay. AI has made that not okay to skip now,” he said. “IT teams have to do two things, which is they have to learn this tooling and how it can make them move faster, and they have to drive it back to basics on maintenance and simplicity, or you just can’t compete in this world.”

Red Hat offers two core tools for exactly that work: Ansible, its IT automation platform that handles patching and configuration at scale, and Red Hat Enterprise Linux, which provides the stable, certified operating foundation that most enterprise workloads run on — including, increasingly, AI, according to Hicks. But the complexity of treating AI as a managed workload — with zero-vulnerability image requirements, rapid build cycles and divergent stability needs across environments — is compounding the challenges IT teams face.

The result is two distinct infrastructure tracks that must coexist within the same enterprise: AI workloads demanding ultra-fast update cycles to stay secure, and mission-critical systems where IT teams are too nervous to touch anything, even when they should. That extends to Common Vulnerabilities and Exposures images, which are system snapshots scrubbed of known security flaws — easy to demand, but costly to deploy on systems organizations are afraid to update.

“We probably see [the tension] the most in RHEL,” he said. “We have the ubiquitous pull at this point of, ‘Give us zero CVE images,’ which we do. But [customers say], ‘No way, I need that on systems that don’t change, ever.'”

Beyond infrastructure, the shift to agentic AI is fundamentally rewriting what it means to work in almost any field, Hicks argued. For engineers, the craft of writing code is giving way to a new discipline of shaping AI systems to solve problems. For everyone else — legal, accounting, sales — the barriers to building and changing things have effectively collapsed, he added.

“The people that will be the most successful with AI are the people that have the most questions,” Hicks said. “People that know business, they know their domain, those are the ones [that] have no barriers to what they can do and change. It’s just about getting started.”

Stay tuned for the complete video interview, part of SiliconANGLE’s and theCUBE’s coverage of Red Hat Summit 2026:

(* Disclosure: Red Hat sponsored this segment of theCUBE. Neither Red Hat nor other sponsors have editorial control over content on theCUBE or SiliconANGLE.)

Photo: SiliconANGLE