The short version

Red Hat unveiled OpenShift 5, its first major upgrade in seven years, promising in‑place upgrades and AI‑driven troubleshooting. The release adds stronger security, automated GPU allocation and an AI upgrade planner.

What happened

Red Hat announced OpenShift 5 at KubeCon + CloudNativeCon and said it can be upgraded from OpenShift 4 without reinstalling the platform.

The new version includes an AI‑powered troubleshooting assistant called Lightspeed, which can investigate cluster events, suggest fixes and produce visual performance charts from natural‑language queries.

Security improvements feature hardened multitenancy, automated certificate management, network isolation and post‑quantum cryptography for control‑plane traffic.

OpenShift 5 also introduces automated GPU resource allocation for AI inference and traditional workloads, together with an AI‑driven upgrade planner that checks application dependencies before updates.

The background

OpenShift 5 marks the first major version change in seven years, reflecting Red Hat’s push to embed AI across its hybrid‑cloud stack.

The company highlighted growth in its virtualization business, noting OpenShift Virtualization now generates about $800 million in revenue.

What it means for owners and operators

Owners can avoid costly reinstall projects and reduce downtime because upgrades can be applied in place.

Operators gain an AI assistant that surfaces performance data and suggests remedial actions, cutting troubleshooting time and reliance on specialised staff.

Stronger security controls help meet compliance on multitenant clouds, especially where regulators demand strict isolation.

Automated GPU allocation can improve utilisation of expensive AI hardware, potentially lowering compute spend.

Should I plan a full migration to OpenShift 5 now?
If you are already on OpenShift 4, you can test the in‑place upgrade in a non‑production environment to assess any dependency issues before committing.
Do I need new hardware for the AI features?
The AI‑driven upgrade planner works on existing clusters; GPU automation only benefits environments that already have GPU resources.

The numbers

OpenShift Virtualization generated $800 million in revenue, according to Red Hat.

NASA migrated roughly 2,400 virtual machines and Gates Corp. moved more than 1,300, reporting a 50 % boost in compute capacity.

A Deloitte survey cited in related reporting found only 21 % of organisations have a mature agentic AI governance model.

Security research notes a 32 % rise in malicious prompt‑injection attempts between November 2025 and February 2026, with 22 distinct payload techniques observed.

$800m

OpenShift Virtualization revenue

2,400

NASA virtual machines migrated

1,300

Gates Corp. virtual machines migrated

32%

Rise in malicious prompt‑injection

Source: as reported by siliconangle.com.

What to do in your business tomorrow

  1. Review your GPU utilisation reports and enable the automated allocation feature if you have under‑used AI hardware.
  2. Pilot the in‑place upgrade on a non‑production cluster to spot any dependency issues before a full rollout.
  3. Check your current security settings against compliance requirements for multitenant environments and tighten isolation where needed.

Related coverage

Sources

  • siliconangle.com: Red Hat adds agentic AI operations, stronger security in OpenShift 5 and RHEL updates
  • MSN: Red Hat and NVIDIA Built AI Agent Security Controls Agents Cannot Override
  • Business Wire: Red Hat AI Factory with NVIDIA Accelerates Agentic AI Security Capabilities Across the Open Hybrid Cloud
  • HPCwire: Red Hat AI 3.5 Adds Safety, Observability and Multi-Tenant GPU Controls
  • Tech Times: Red Hat and NVIDIA Built AI Agent Security Controls Agents Cannot Override