Red Hat expands OpenShift for AI agents using Nvidia OpenShell

Red Hat is expanding its cloud-native Kubernetes infrastructure to support production AI agents by addressing security, resource management, and inference demands. The company is integrating Nvidia's OpenShell runtime into its enterprise AI platform for agent sandboxing and least-privilege controls, while participating in projects like llm-d to handle multi-turn inference workloads. These efforts aim to provide enforceable security boundaries and efficient orchestration as enterprises adopt autonomous agents. Open article →