Red Hat released OpenShift 4.20, the latest version of its hybrid cloud application platform, adding new capabilities for AI workloads, tighter security controls, and expanded virtualization support across on-premises and sovereign cloud environments.

OpenShift is a Kubernetes-based application platform that helps organizations build, deploy, and manage containers across hybrid and multi-cloud environments. It provides a foundation for running both traditional and cloud-native applications with integrated tools for security, automation, and DevOps.

With version 4.20, Red Hat is emphasizing three core themes: security hardening, AI acceleration, and virtualization flexibility.

Strengthened platform security

OpenShift 4.20 introduces several new features aimed at enterprise security and compliance. The release includes initial support for post-quantum cryptography (PQC) algorithms for mutual TLS (mTLS) traffic. It also expands Red Hat’s zero-trust approach with bring-your-own OpenID Connect for identity management and External Secrets Operator (ESO) for lifecycle control of secrets sourced from external management systems.

For OpenShift Platform Plus customers, Red Hat is making Advanced Cluster Security 4.9 generally available, along with enhancements to the Trusted Artifact Signer and Trusted Profile Analyzer. A zero-trust workload identity manager – designed to provide identity attestation for both machines and human users – is expected later this year.

AI workloads move toward production

As enterprises look to operationalize AI, OpenShift 4.20 introduces new orchestration and data-handling tools. The LeaderWorkerSet (LWS) API is designed to simplify the management of distributed AI workloads, while the Image Volume Source feature offers faster deployment of AI models without rebuilding containers. Combined with Red Hat OpenShift AI, these capabilities aim to reduce the time it takes to move AI projects from experimentation to production.

Virtualization & hybrid cloud flexibility

The release also expands OpenShift Virtualization to manage virtual machines (VMs) alongside container workloads, adding CPU load-aware rebalancing and Arm architecture support. Support for Oracle Cloud bare-metal deployments gives organizations more control over workload placement and data residency.

Red Hat says OpenShift 4.20 is now generally available and can be deployed across data centers, public clouds, and edge environments.