Brievox

NVIDIA launches open platform for securing autonomous AI agents

NVIDIA announced an open software platform and reference system design for adding safeguards to autonomous AI agents. The described design combines a host-CPU runtime with an out-of-band watchdog on a DPU; one source characterizes it as continuous in-silicon monitoring, while the other describes the watchdog as something an agent cannot see.

2 sources Published Updated

Image: StorageReview

What we know

Reported by one source

  • NVIDIA launched an open software platform and reference system design intended to put guardrails around autonomous AI agents. — StorageReview
  • The design uses a secure runtime on the host CPU and an out-of-band watchdog on a DPU; the agent cannot see the watchdog. — StorageReview
  • The platform is presented as a reference for continuous in-silicon agent monitoring. — NVIDIA Developer Blog
  • NVIDIA says the platform has more than 100 partners, with Anthropic and SpaceXAI among those named. — StorageReview

Open questions

  • The sources do not independently establish the platform's effectiveness, deployment status, or the specific security incidents cited as motivation. — StorageReview, NVIDIA Developer Blog

Coverage 2 publishers

  1. NVIDIA Developer Blog

    NVIDIA Open Agent Safety Platform: A Reference for Continuous In-Silicon Agent Monitoring

  2. StorageReview

    NVIDIA Open Agent Safety Platform: OpenShell on the CPU, Sentry on BlueField-4, and 100-Plus Partners From Anthropic to SpaceXAI

Articles stay on their publishers’ sites; each link opens the original.