Enterprise AI infrastructure modernization is at a critical crossroads. Organizations face intense pressure to deploy AI while grappling with decades of technical debt. According to Red Hat CEO Matt Hicks, returning to IT fundamentals is now the only viable path forward.
"We have a feeling of, 'We have to harness what this technology can do,'" Hicks said. "IT teams have been dealing with tons of technologies but not always maintaining technical debt. That is all crashing down on them."
Speaking at Red Hat Summit 2026, Hicks explained that the race to deploy AI is won or lost long before the first model hits production. Boards demand AI returns, but IT teams are buried under unresolved complexity.
The solution, Hicks argues, is a "back-to-basics" approach. IT teams must learn to patch systems and maintain simplicity. AI has made skipping maintenance no longer acceptable.
Red Hat offers two core tools for this work: Ansible for automation and Red Hat Enterprise Linux for a stable foundation. But the complexity of running AI workloads-with zero-vulnerability image requirements-is compounding challenges.
Two distinct infrastructure tracks must coexist: AI workloads needing ultra-fast updates, and mission-critical systems where IT teams are afraid to touch anything. "We see the tension the most in RHEL," Hicks said. "Customers want zero CVE images, but say, 'No way, I need that on systems that don't change, ever.'"
Beyond infrastructure, Hicks believes agentic AI is rewriting what it means to work in almost any field. For engineers, writing code is giving way to shaping AI systems. For everyone else, the barriers to building and changing things have collapsed.
"The people that will be most successful with AI are those with the most questions," Hicks said. "People who know their business and domain-they have no barriers to what they can do. It's just about getting started."