CrowdStrike and NVIDIA develop secure-by-design AI blueprint

Cybersecurity company CrowdStrike and NVIDIA have unveiled a secure-by-design AI blueprint for the operation of autonomous AI agents both locally and in the cloud.

The blueprint involves integrating protection from the CrowdStrike Falcon platform into the open-source NVIDIA OpenShell run-time, which enforces policy-based guardrails for the deployment of autonomous agents. The architecture is designed to integrate natively into the AI agent stack to enable organisations to operationalise autonomous systems while preserving governance, visibility and control throughout the development and operational lifecycle.

The OpenShell runtime provides isolated sandboxes with private inference and built-in policy enforcement. Falcon can extend protection to local agents running on NVIDIA DGX Spark or NVIDIA DGX Station hardware, and can extend this security to agents in the cloud built on the NVIDIA AI-Q Blueprint.

Falcon will secure every prompt, response and agent action in real time, while enforcing host-level controls and continuous behavioural monitoring across system activity and agent execution.

CrowdStrike Chief Business Officer Daniel Bernard said such capabilities will be important as AI technology evolves.

“As we enter the agentic era, agents no longer simply assist — they act,” he said. “This shift fundamentally changes the security equation, and security must be embedded into the AI stack itself. Together with NVIDIA, we are delivering a secure-by-design architecture that enables organisations to operationalise agents with confidence and control.”

NVIDIA VP for Enterprise Platforms Justin Boitano added that integrating Falcon with the NVIDIA Agent Toolkit will enable enterprises to “build and scale safer, autonomous AI agents to help transform their operations, empower every employee, and securely generate intelligence at the speed of business.”

Image credit: iStock.com/BlackJack3D