NVIDIA launched the Open Agent Safety Platform, an open platform designed to secure AI agents from testing through deployment. It combines OpenShell, which creates software boundaries and enforces policies, with Sentry, a hardware-based watchdog running on NVIDIA BlueField-4 DPUs that can monitor and quarantine agents within milliseconds.
The platform aims to provide full-stack security, controlling agents across software, computing infrastructure, and robotics. NVIDIA says more than 100 organizations are working with the technology, including Anthropic, Microsoft, Cisco, CrowdStrike, Salesforce, SAP, and SpaceXAI.
The initiative reflects growing concerns that AI agents can bypass application-level security controls while carrying out tasks. NVIDIA argues that stronger safeguards must therefore operate outside the AI model itself, with continuous monitoring, access controls, auditing, and human oversight.
