NVIDIA OpenShell Brings Sandboxed Runtime to Secure AI Agents
NVIDIA has introduced OpenShell, a secure runtime component of its Open Agent Safety Platform, offering sandboxed execution and policy-based access controls for enterprise AI agents, with Hyundai and AIM Intelligence among early adopters. For defenders, this closes a meaningful gap in agentic AI governance — specifically the lack of runtime-level controls over what autonomous agents can access, execute, and exfiltrate within enterprise environments. Realising the full benefit will depend on integration maturity, breadth of enterprise tool coverage, and how consistently policy frameworks are configured and maintained across deployments.