Red Hat integrates agentic AI and post-quantum security into OpenShift and RHEL

The rise of agentic AI

Red Hat is evolving its AI capabilities from simple assistants into 'agentic' systems. While previous tools might just answer questions, these new agents can actively investigate operational problems and recommend specific corrective actions. This shift is designed to help administrators manage complex hybrid cloud environments more efficiently.

The OpenShift Lightspeed assistant now allows users to interact with system data using natural language. For instance, an administrator can request a visualization of CPU usage over a specific week and ask the system to overlay application errors on top of that data. The platform then generates these charts, which can be saved to continuously updating dashboards.

When a system failure occurs, the agentic troubleshooting feature takes over by examining related cluster events. It reasons through the possible causes of the breakdown and presents a plan for remediation. However, the system does not act on its own; humans must still approve any suggested actions before they are executed.

Simplifying infrastructure upgrades

One of the primary goals of OpenShift 5 is to make the transition between software versions less painful for large organizations. Red Hat notes that while upgrading the underlying Kubernetes software is relatively straightforward, managing the various applications and dependencies running on top of it is often the difficult part. To solve this, they introduced an AI-driven upgrade planner.

This planner scans the entire environment to identify potential compatibility problems before any changes are made. It looks at application dependencies, Kubernetes application programming interfaces, and optional operators to flag risks. This proactive approach aims to reduce the downtime and errors typically associated with major platform transitions.

Unlike previous versions, OpenShift 5 can be installed as an in-place update from OpenShift 4. This means companies do not have to reinstall their software or migrate their existing workloads to a new environment. Red Hat describes this as a simple upgrade that provides more predictability for companies running long-lived enterprise applications.

Hardening OS security

The upcoming releases of Red Hat Enterprise Linux (RHEL) 10.3 and 9.9 focus heavily on integrity and future-proofing. A major addition is the support for sealed images, which use customer-controlled cryptographic keys to verify the operating system. This process checks the integrity of the system starting from the bootloader, moving through the kernel, and ending at the root file system.

By using sealed images, organizations can prevent unauthorized modifications to the OS and block compromised software components from running. This level of security is particularly important for regulated organizations and sovereign cloud providers. These entities often require strict separation between different tenants and must prove they meet specific compliance standards.

Additionally, Red Hat is incorporating post-quantum cryptography into its ecosystem. This technology is designed to protect communications against the potential power of future quantum computers. The updates include support for the post-quantum Module-Lattice-Based Key-Encapsulation Mechanism encryption algorithm for applications built using the Go programming language.

Local AI and developer tools

Red Hat is also providing ways for organizations to run AI inference locally. They are introducing a self-managed version of the RHEL command-line assistant, which is currently available in developer preview. This is specifically designed for air-gapped or highly regulated environments where sending operational data outside the local network is prohibited.

To support these agents, Red Hat is implementing a Model Context Protocol (MCP) server. This allows compatible AI assistants to investigate live Linux systems while maintaining strict access controls. It also integrates with the Performance Co-Pilot, an open-source toolkit that helps agents analyze historical performance data alongside current conditions to find the root cause of problems.

On the development side, the Developer Hub 2.1 update provides a framework for AI coding agents. It gives these agents governed access to approved software components and standardized workflows. This ensures that when AI helps write code, it follows the established procedures and security rules of the organization.