NVIDIA Moves AI Agent Safety Controls Outside the Agent with OpenShell and BlueField-4 Sentry
shashib · x · 2026-09-29
NVIDIA's new Open Agent Safety platform (announced Sept 28) runs the controls that stop an AI agent on separate hardware and software the agent can't reach. OpenShell, open-source, logs every agent action and enforces rules on Vera processors; Sentry is a watchdog blueprint running on dedicated BlueField-4 network chips that halts misbehaving agents in milliseconds.
The motivation: every incident NVIDIA cites follows one pattern—agents bypassing in-app safety controls to finish their task. Over 100 organizations work with the platform technologies, 120+ belong to the Open Secure AI Alliance, and the launch summary names 18 partner companies. It applies the classic infra principle of separating management planes from workloads to agent governance.
Related event: NVIDIA Launches Open Agent Safety Platform with 100+ Partners(35 posts)→
More from coding & agent
- Baseten's Guide: Evaluate Open Models for Coding Agents on Your Own PRs — baseten · 2026-11-16
- Google Research open-sources PaperOrchestra, a multi-agent framework for writing AI papers — google-research · 2026-11-16
- ffmpeg-skill gives coding agents local video editing via 42 FFmpeg tools and MCP — kajisho5 · 2026-11-16
- Agents Now Reply to Slack Messages on Their Own, Resolving a Prod Incident Ghost-to-Ghost — iamrobotbear · 2026-11-16
- Engineer's Interview Prep Method: Use Claude to Find Which Answers Undersell Your AI Work — brandon_galang · 2026-11-16
- DeepSeek open-sources its app frontend "chat-demo", already at 3.9k stars — NielsRogge · 2026-11-16