Nvidia Releases Software It Says Can Prevent AI Agents From Going Rogue

The chip giant said its platform allows developers to continuously monitor and govern AI behavior.

Read the full story at Wall Street Journal Technology ↗

Timeline · 1 report

2026-09-28 12:00 · Wall Street Journal Technology
Nvidia Releases Software It Says Can Prevent AI Agents From Going Rogue

More stories

NVIDIA Open Agent Safety Platform: A Reference for Continuous In-Silicon Agent Monitoring — NVIDIA Technical Blog
Nvidia’s Answer to Rogue Agents Is an Open-Source AI Security System — Wired AI
Add Runtime Controls to AI Agents with NVIDIA OpenShell — NVIDIA Technical Blog
How NVIDIA DSX MaxLPS Maximizes AI Factory Throughput and Efficiency — NVIDIA Technical Blog
Nvidia (NVDA) Boosts Stock Buyback Authorization by $150 Billion — Bloomberg AI
Artificial Analysis launches the Cyber Index Alliance with partners Collinear, IBM, Nvidia, and Vercel to evaluate how AI agents find and fix vulnerabilities (Artificial Analysis) — Techmeme
Nvidia launches record $150bn share buyback — Financial Times AI
Nvidia unveils new system to put guardrails on AI agents — The Hill Technology

Get the daily brief of stories like this at 6:30 every morning →