TL;DR: CNCF launched the Certified Kubernetes AI Conformance Program to standardize and confirm compatibility for running AI workloads on Kubernetes. The program is already available, v1.0 is certified, and v2.0 is planned for next year.

• Why: reduce fragmentation and give AI portability and reliability across different environments.
• What’s included: minimum requirements and configurations for popular AI/ML frameworks on Kubernetes; compatibility tests.
• Stats: 82% of companies build custom AI; 58% run it on Kubernetes; 90% consider open source critical.
• Status: announced Nov 11, 2025 at KubeCon + CloudNativeCon North America (Atlanta); v1.0 has already certified the first participants; v2.0 roadmap is for next year.
• Who supported / got certified: AWS (EKS), Google Cloud (GKE), Microsoft Azure, Broadcom (VKS), Red Hat (OpenShift / OpenShift AI), Oracle OCI, CoreWeave, Akamai, Giant Swarm, Kubermatic, Sidero Labs.
• How it’s governed: open development on GitHub; an AI Conformance working group with an open charter.
• Key test areas: GPU integration, volumes, networking at the job level.
• Base: builds on the successful Certified Kubernetes Conformance Program (100+ compatible systems). The goal is portability, reproducibility, and no vendor lock‑in.

And now there’s the first implementation from VMware.