Kubernetes cements role as AI & cloud native backbone

Kubernetes use in production has reached 82% among container users, according to new survey data from the Cloud Native Computing Foundation, which also points to wider use of Kubernetes for generative AI workloads.

The CNCF said 66% of organisations that host generative AI models now use Kubernetes for some or all of their inference workloads. The data suggests that Kubernetes has become a default layer for parts of AI processing in production environments.

The survey also found cloud native adoption has moved into mainstream IT operations. It reported that 98% of surveyed organisations have adopted cloud native techniques. It said 59% now run "much" or "nearly all" of their development and deployment in cloud native environments.

The CNCF described a shift in barriers facing teams as cloud native practices spread. It reported that cultural and organisational change now sits ahead of technical factors as the most cited challenge.

"Over the past decade, Kubernetes has become the foundation of modern infrastructure," said Jonathan Bryce, Executive Director, CNCF. "Now, as AI and cloud native converge, we're entering a new chapter. Kubernetes isn't just scaling applications; it's becoming the platform for intelligent systems. This community has the expertise to shape how AI runs at scale, and we have a massive opportunity to build something open, powerful, and impactful for the next ten years."

Production uptake

The survey data suggests a sustained rise in production use of Kubernetes. The CNCF said 82% of container users now run Kubernetes in production, up from 66% in 2023.

The organisation also reported signs of a slowing adoption curve. It noted 10% of organisations remain in the early stages of cloud native adoption or do not use cloud native approaches.

For technology leaders, the figures frame Kubernetes less as a specialist tool for platform teams and more as a standard operating component of modern application delivery. The survey data also indicates that cloud native tooling now sits inside routine engineering processes for many organisations, rather than pilot programmes.

AI workloads

The CNCF positioned Kubernetes as a common platform choice for running inference workloads at scale. It reported that two-thirds of organisations hosting generative AI models rely on Kubernetes for at least part of that processing.

The data also points to a gap between infrastructure readiness and model deployment frequency. The CNCF said only 7% of organisations deploy models daily. It said 47% deploy models occasionally.

The survey indicates many organisations still sit at an earlier stage of operational AI maturity. The CNCF said 44% of respondents do not yet run AI or machine learning workloads on Kubernetes.

GitOps adoption

The survey also separated organisations by levels of operational maturity and linked that maturity to specific approaches in software delivery. The CNCF said 58% of "cloud native innovators" use GitOps principles extensively, compared with 23% of "adopters".

The findings also highlighted internal developer platforms as an active area of engineering investment. The CNCF cited the Backstage project, which it said ranks as the fifth CNCF project by velocity. Backstage focuses on internal developer portals.

The survey framed these tools as part of a standardisation push. Many organisations now seek tighter governance of software delivery patterns across teams. GitOps workflows and developer portals often sit at the centre of that standardisation effort.

Observability focus

The survey highlighted observability as another key area of activity. The CNCF said OpenTelemetry is now the second-highest-velocity CNCF project. It reported more than 24,000 contributors.

It also cited increased use of profiling in observability tooling. The CNCF said nearly 20% of respondents now use profiling as part of their observability stack.

The data reflects continued investment in monitoring and diagnostics as Kubernetes footprints expand. Larger environments often increase demand for standardised telemetry data and a clearer view of software performance across distributed systems.

Cultural barriers

The survey found that organisational issues now outrank training, security, and complexity as a barrier to cloud native adoption. The CNCF said "Cultural changes with the development team" was the top challenge, cited by 47% of respondents.

Other factors still ranked highly, but below cultural change. The CNCF said lack of training and security each received 36%, while complexity received 34%.

"Enterprises are aligning around Kubernetes because it has proven to be the most effective and reliable platform for deploying modern, production-grade systems at scale-including AI-and because of the ecosystem and community that support it," said Hilary Carter, Senior Vice President of research at Linux Foundation Research. "This year's data shows that the next phase of cloud native evolution will be as much about people and platforms as it is about the tech itself. Organisations that invest in both will have a clear advantage."