About a decade ago, containers were a science experiment. Today, they run virtually every new application built, and the next wave of AI software is accelerating the shift faster than anyone predicted.
The shift carries profound implications for IT decision makers navigating infrastructure choices. As enterprises race to deploy AI applications, they face a dual reality: new software runs in containers while decades of legacy applications stay in virtual machines. The organizations that figure out how to manage both on one platform will move fastest.
"Essentially, all new applications are written to run on Kubernetes," said Dan Ciruli, vice president and general manager of cloud native technologies at Nutanix.
Ciruli spent more than a decade in the open source community, co-founding the OpenAPI Initiative and serving on the Istio Steering Committee before joining Nutanix, where he now oversees cloud native strategy. His career includes seven years at Google Cloud, where he watched Kubernetes grow from an internal Google project into the dominant force in enterprise software.
The shift is already measurable. According to the 2026 Nutanix Enterprise Cloud Index (ECI), a survey of 1,600 cloud, IT and engineering executives from across the globe, 87% of organizations expect application containerization to increase over the next three years. The same report found that 85% of executives identify AI as a factor accelerating their adoption of container, while 83% say they are already building new applications in containers.
The data reveals a hybrid reality. The ECI found that 71% of organizations run their AI-enabled applications on a mix of traditional apps in virtual machines and modern apps in containers. Meanwhile, 82% of executives say their on-premises infrastructure is not fully ready to support AI workloads, a gap that puts pressure on IT teams to modernize without disrupting existing systems. The research, conducted by Wakefield Research, spans companies with 500 or more employees in markets including the United States, United Kingdom, Germany, Japan and India.
Containers are lightweight, portable software packages that bundle an application with its dependencies so it runs consistently across any environment. Kubernetes is an open source orchestration platform that automates the deployment, scaling and management of containerized applications across data centers, clouds and edge devices.
Why Have Containers and Kubernetes Become the Default?
Ciruli traces the technology's evolution from a science experiment eight years ago to a viable option three or four years back to today's status as the default choice for application development.
"It is safe to say at this point that containers in general and Kubernetes in particular are the de facto standard for developing and deploying new applications," Ciruli said.
All new applications are being built as cloud-native applications, and that means containers and Kubernetes. Agentic AI applications are no different. They are new applications, so they are being built to run in containers from the start. Kubernetes becomes the default deployment model.
The rapid emergence of agentic AI applications exemplifies the trend.
"The word agentic is only about 18 months old," Ciruli said. “All of these agents are new applications. They're all being written to be deployed in containers.”
Will Virtual Machines Disappear?
The narrative that virtualization is no longer necessary is one of the most misleading stories circulating in the market today, Goel explained in his blog post The Race to Fill the Virtualization Gap. He see reality is far more nuanced.
"Changes in the virtualization market have created uncertainty and pushed many organizations to reevaluate their infrastructure strategies," explained Goel. "The operational responsibility does not vanish when you remove the virtualization layer. It just shifts back onto infrastructure teams that are already stretched thin."
For many enterprise workloads, virtualization delivers more operational value than organizations would gain by standardizing entirely on bare metal, Goel explained. He believes infrastructure decisions are rarely based on performance alone. Operational agility, security, resilience, scalability and manageability all play equally important roles.
"Containers and virtualization solve different problems," Goel said. "Containers focus on application packaging and portability. Virtualization focuses on infrastructure abstraction, workload isolation and operational management."
"Organizations increasingly need platforms that can run virtual machines and containers side by side, provide centralized visibility across environments, simplify lifecycle and infrastructure management, deliver consistent security and governance policies, and support application portability across data centers, cloud and edge," Goel said.
"The future is not about choosing one model. It is about managing everything at once."
Rewriting legacy applications for containers rarely pays off, leaving IT teams to manage both environments simultaneously for the foreseeable future, explained Ciruli.
"At essentially every enterprise, they have decades worth of applications running in virtual machines that are for the most part going to stay in virtual machines," Ciruli said. "Refactoring existing applications to run as cloud-native isn't always worth the effort if they're already doing their job. Many of those workloads will stay in VMs for years."
The ECI data supports this hybrid reality. The report found that 71% of organizations run AI-enabled applications on a mix of traditional apps in VMs and modern apps in containers, while 14% run their AI-enabled apps directly on bare metal servers. This creates operational complexity as teams centralize infrastructure management across both paradigms.
What Operational Challenges Do IT Teams Face?
The transition to container-based infrastructure introduces new skill requirements that present challenges for IT organizations.
"It does take additional knowledge to run containerized applications, to run Kubernetes," Ciruli said.
The challenge compounds when teams manage containers and VMs with separate tools, policies and processes. Ciruli advocates for converged platforms that eliminate operational silos.
"Companies that combine these on one platform allow themselves to run essentially any application, it doesn't matter if it's virtualized or containerized," he said.
This approach enables consistent security policies, backup procedures, disaster recovery protocols and networking configurations across both environments.
Vendors approach the same problem from different angles. What sets Nutanix Kubernetes Platform apart, Ciruli said, is that it brings together three things that are usually fragmented: enterprise-grade distributed storage, a proven virtualization layer and a full Kubernetes platform. The key principle is co-locating compute with data, which becomes even more critical with AI and GPUs. In a distributed environment, the platform places workloads exactly where the data lives, reducing latency and improving performance.
The same principle applies at the edge, where organizations need to bring the application to the data source. That means running containers or VMs directly at the edge with the same operational model used in the data center or cloud.
How Will AI Change Kubernetes Management?
Ciruli sees AI playing a significant role in simplifying Kubernetes operations.
"How do we make it easier for the average operator to run a Kubernetes cluster? How do we make the tools smart so that the tools are doing more of the groundwork?" he said.
He compares the goal to how modern cars continuously tune themselves rather than requiring owners to be mechanics.
"AI has the potential to significantly simplify operations," Ciruli said. "We're moving toward a model where the system can understand intent, expressed in natural language, and translate that into the configurations needed to run an application. More importantly, it can observe and diagnose itself by correlating signals across the environment."
Over time, he believes that intelligence will reduce the operational burden and make these systems much easier to manage. It’s important to make sure developers can build and run AI applications consistently, whether in the cloud, on-premises or at the edge, without rethinking the underlying platform each time.
What Comes Next?
Looking ahead, Ciruli sees AI becoming deeply embedded across all enterprise applications rather than remaining a separate category.
"You want AI everywhere," he said, whether in business process workflows, email systems or sales data. This vision extends to infrastructure itself, where he expects organizations will want to run containerized and virtualized applications together on AI-enabled platforms.
For Ciruli, the convergence of these technology worlds represents an exciting inflection point. As enterprises navigate this transition, he emphasizes the importance of standardization to avoid fragmentation as different teams lean into new technologies in different directions.
The path forward requires balancing innovation with operational reality: embracing containers and AI for new development while maintaining robust support for existing virtualized workloads, all while making these increasingly complex systems easier for IT teams to manage.
"I'm firmly convinced that in three years we won't differentiate between applications and AI applications," Ciruli said. "We'll just call them applications."
Related stories:
Ken Kaplan is Editor in Chief for The Forecast by Nutanix. Find him on X @kenekaplan and LinkedIn.
© 2026 Nutanix, Inc. All rights reserved. For additional information and important legal disclaimers, please go here.
Related Articles
