Nvidia launches Open Agent Safety Platform with OpenShell and Sentry for AI agent guardrails
Nvidia has launched its Open Agent Safety Platform, an open-source and reference system for AI agent safety built around OpenShell software controls and Sentry hardware monitoring.
TLDR
Nvidia on Sept. 28 introduced the Open Agent Safety Platform, a new AI agent safety stack built around two components: OpenShell and Sentry. OpenShell is open-source software that sits between an AI agent and enterprise systems, letting operators define and enforce limits on access to files, credentials, tools, processes and networks. Sentry adds a hardware-level monitoring and enforcement layer on Nvidia infrastructure, which the company says can watch agent behavior in real time and quarantine suspicious agents quickly. Nvidia is positioning the system as a shared framework spanning labs, enterprises and hardware providers, with partners including Cisco, Microsoft, Oracle, CoreWeave, Dell, HPE, Lenovo, ARM and Intel. CNBC also reported Nvidia is working with Anthropic to integrate cloud managed agents with OpenShell.
Combined views
2.4M
24 Sources, first seen 6h ago

