Red Hat and Nvidia expand partnership to secure autonomous AI agents
At Nvidia's GTC event in Berlin, Red Hat and Nvidia expanded their collaboration to help enterprises secure autonomous AI agents, govern access to AI models and deploy workloads on Nvidia's upcoming Vera Rubin data center infrastructure. The updates center on Red Hat AI Factory with Nvidia, combining agent runtime controls, model inference software and hardware support across hybrid clouds, and Red Hat said it is working to bring Nvidia's OpenShell sandbox runtime into its AI platforms as a native capability. The companies aim to place security controls outside the agent and, where appropriate, outside the host operating system, pairing OpenShell with Nvidia's Sentry framework on BlueField data processing units, where DOCA Argus monitors workloads from a separate hardware security domain. A joint reference design adds Model-as-a-Service access controls and an AI Gateway alongside vLLM, llm-d, Nvidia Dynamo and NIM, and Red Hat said its Enterprise Linux for Nvidia now supports Vera CPUs and is prepared for Vera Rubin.
Security controls that sit outside the agent and host operating system are being positioned as the basis for letting enterprises run autonomous agents, per Nvidia's John Fanelli, who said an agent should not enforce its own boundaries.
Red Hat Inc. today expanded its collaboration with Nvidia Corp. to help enterprises secure autonomous artificial intelligence agents, govern access to AI models and deploy workloads on Nvidia’s upcoming Vera Rubin data center infrastructure.
Announced at Nvidia’s GTC event in Berlin, the updates center on Red Hat AI Factory with Nvidia, combining agent runtime controls, model inference software and hardware support across hybrid cloud environments. Red Hat also said it’s working to integrate Nvidia’s OpenShell Reference Architecture into its AI platforms as a native capability.
The security approach addresses two risks: agents exceeding their authorized boundaries and attackers compromising agents through techniques such as prompt injection or supply chain attacks. The companies aim to place security controls outside the agent and, where appropriate, the host operating system as well.
“An agent shouldn’t enforce its own boundaries,” said John Fanelli, vice president of enterprise software at Nvidia, in a statement. “Those controls belong outside the agent, in the runtime and the infrastructure beneath it.”
OpenShell provides a sandbox runtime for agents. Beneath that software layer, Nvidia’s Sentry security framework runs on the company’s BlueField data processing units, using its DOCA Argus software to monitor workloads from a separate hardware security domain.
#Red Hat#Nvidia#Red Hat AI Factory#Nvidia OpenShell#Nvidia BlueField#Nvidia Vera Rubin#Red Hat OpenShift#Red Hat Enterprise Linux for Nvidia#Nvidia Nemotron 3.5 Super VL#vLLM#Dell Technologies#Lenovo