A Win for Defenders: CrowdStrike and NVIDIA Extend Security Across the AI Stack

CrowdStrike and NVIDIA are collaborating on the NVIDIA Open Agent Safety Platform, an open reference design to extend security deeper into the agentic stack and establish stronger boundaries for autonomous AI.

AI agents are already operating across the enterprise. As they become more autonomous in accessing data, using credentials, executing code, and acting across critical systems, the security boundary is changing with them.

The question is no longer whether enterprises will adopt agents. It's how they give agents greater autonomy without giving up control.

Today, NVIDIA introduced the NVIDIA Open Agent Safety Platform, an open source platform and reference design that brings multi-layered security controls across the software, runtime, and infrastructure where agents operate. CrowdStrike and NVIDIA are collaborating on this effort to extend our work together deeper into the agentic stack.

Model safeguards are essential, but they aren't the entire security boundary. When agents operate with identities, credentials, and access to critical systems, enterprises need controls that work where agents run and act.

Agent safety is a cybersecurity problem. It demands a full-stack approach.

That's the approach CrowdStrike and NVIDIA are advancing together. We are empowering AI to transform cyber defense and securing the AI transforming the enterprise.

The Agent Can't Be Its Own Security Boundary

An agent shouldn't be responsible for enforcing its own security boundaries.

We already understand this everywhere else in cybersecurity. We don't rely on applications or workloads to determine whether they've been compromised. We establish independent controls around them to govern access, monitor behavior, and enforce policy.