Nvidia Launches Open Agent Safety Platform for Secure AI Deployment

Nvidia has launched the Open Agent Safety Platform, an open-source software platform and reference system design aimed at providing security controls for AI agents from testing to deployment. The platform consists of two primary components: Nvidia OpenShell and Nvidia Sentry.

Nvidia OpenShell is a secure runtime boundary for AI agents that traces agent actions and enforces defined policies while agents operate on Nvidia Vera CPUs. It can also be extended to work with third-party computing platforms, including systems based on technology from Arm and Intel.

Nvidia Sentry continuously monitors agent behavior and can quarantine agents within milliseconds if they attempt to operate outside defined boundaries. The platform has been developed in collaboration with over 100 organizations across various industries, including technology, financial services, and robotics.

Companies such as Salesforce and SAP are already integrating OpenShell into their products, enabling teams to monitor agent activity and approve or reject permission requests within the application. Nvidia said that OpenShell software and related tools are available through its developer resources and GitHub.