When Broadcom’s comprehensive blind survey of IT executives called Private Cloud Outlook 2026 was published, one finding stood out above all others: enterprises are not behind on AI implementation because they lack ambition. They are behind because their infrastructure and operating models have not kept pace with the actual demands of AI.
That finding became the basis for this special report. Today, we are publishing From Modernization to AI Readiness: The New Platform Imperative, with a focus on application modernization, private cloud maturity, and the rise of platform engineering as a core IT function.
The Modernization Change Is Real, and AI Is Widening It
Fifty-seven percent of enterprise IT teams surveyed said that their top modernization approach right now is adding AI capabilities to existing applications. That is the most-selected response in our survey, ahead of rehosting, replatforming, and refactoring.
That is a significant shift. It means enterprises are shifting their views of what constitutes modernizing in the age of AI. All approaches still exist, however a new approach is taking shape where adding AI capabilities to existing applications is trending as the new modern.
The challenge is that most application portfolios are not ready for that. Seventy-two percent of enterprises have modernized less than half their application portfolio. Seven in ten IT organizations are trying to add AI capabilities to a foundation that is still largely unmodernized. That creates real pressure.
The old model of sequential transformation, where modernization comes first and AI comes second, no longer fits the operating reality most IT teams are navigating. Enterprises need a platform that supports both traditional and AI-enhanced workloads in parallel. That is not a future architectural goal. It is a current operational requirement.
Private Cloud Has Matured for This Moment
One of the most important shifts in this year's research is how enterprise IT leaders now view private cloud. The data reflects a clear confidence: 93% agree that private cloud delivers the reliability business-critical applications demand, and 92% say it provides the financial transparency and predictable costs needed to govern AI infrastructure spend. For organizations operating under data residency or sovereignty requirements, four out of five say their private cloud mostly or fully enables deployment in those constrained environments.
That confidence is grounded in capability. Private cloud has matured across dimensions that matter most for AI: unified workload support for containers, VMs, and AI workloads running together under consistent policy; automated compliance guardrails built into the platform rather than added after deployment; self-service provisioning that lets development teams move without waiting on manual infrastructure processes; cost visibility and predictability for AI infrastructure planning; and persona-based services that give developers, security teams, FinOps leaders, and ML engineers role-specific experiences from a common platform.
The private cloud of three years ago was not positioned to run production AI at scale. The private cloud enterprises are operating today largely is - and it is also better positioned to address the cost, security and compliance needs that AI introduces as well.
Platform Engineering Has Crossed Into the Mainstream
The third finding that stands out in this report is how quickly platform engineering has become standard practice. Eighty percent of enterprises now have a dedicated platform engineering team. Another 15% plan to establish one within 18 months. The question for most organizations is no longer whether to build a platform engineering function, but how to make it effective.
The answer, according to the data, lies in how platform engineering teams relate to IT infrastructure. Only 24% of platform engineering teams operate completely independently from IT infrastructure. The rest have some form of active collaboration in place. That is encouraging. The problem is that only 12% have made that collaboration formal. Shared governance for AI workload placement decisions, joint operating models, and structured skills development are things most organizations know they need. Few have built the structures to deliver them consistently.
The organizations furthest along in platform engineering maturity are better positioned to operationalize AI at scale. They already have the structures to turn infrastructure capabilities into consumable services across teams. That’s the practical value of getting platform engineering right: it gives IT a way to deliver AI readiness at the pace the business requires.
A Practical Roadmap
What emerges from this research is a practical roadmap: modernize existing applications in parallel with AI adoption, strengthen the private cloud platform against the operational capabilities that define AI readiness and platform maturity, and align the platform engineering and infrastructure teams responsible for delivering it.
None of these are long-horizon initiatives. The data shows that most enterprises already have the foundational elements in place. Private cloud is mature. Platform engineering teams exist. Collaboration between infrastructure and platform teams is underway in most organizations. The work now is to formalize and accelerate what is already happening.
That is the platform imperative. Enterprises need infrastructure that can support AI and a mature operating model that can deliver its benefits consistently across teams.
Read the full report to see the complete findings and specific recommendations for Platform Engineering Leaders, CIOs, and Infrastructure VPs.
Download the Special Report: From Modernization to AI Readiness: The New Platform Imperative