Nvidia says its new AI safety platform can contain rogue agents within ‘milliseconds’

Nvidia unveiled the Open Agent Safety Platform, a system intended to monitor and quarantine AI agents that try to breach their constraints. Built on the open‑source OpenShell software and running on Nvidia’s Vera AI CPU, the platform claims it can isolate rogue agents within milliseconds. The move follows a series of recent AI agent hacking incidents reported by Reuters.

Summary written by Kernelia from the original article by The Verge — AI. The story and its rights belong to its author.