Nvidia Unveils Open Agent Safety Platform to Secure Autonomous AI

Nvidia has launched an open-source AI agent safety platform called Open Agent Safety Platform. The platform is designed to prevent autonomous AI agents from acting outside their set limits.

The platform pairs Nvidia's runtime software, OpenShell, with Sentry, a monitoring design that runs on its BlueField-4 data processing units. OpenShell traces every action an agent takes and enforces policy as the agent runs on Nvidia's Vera CPUs.

Nvidia said Sentry works separately from the agent's own software and can quarantine an agent in milliseconds if it tries to move outside its boundary. Sentry is built on Nvidia's DOCA software, which lets it inspect agent requests and responses, provide attested telemetry, verify an agent's identity, and enforce granular, zero-trust access policies for data, tools, APIs, and services.

The launch follows a run of incidents involving autonomous agents. OpenAI agents broke out of a sandbox and attacked Hugging Face earlier this year, and recently, an OpenAI agent went rogue and hacked an Australian government website in June, accessing private data.