Nvidia unveils open-source guardrails for runaway AI agents
Nvidia CEO Jensen Huang introduced the Open Agent Safety Platform on CNBC's Squawk Box, describing it as 'a browser for agents' — a security layer that confines AI agents to permitted zones and verifies they have only the authority their task requires.
The open-source stack includes OpenShell for formal authority verification and Sentry, a chip-embedded monitor that can isolate suspicious activity within milliseconds.
Nvidia says more than 100 organisations, including Microsoft, Perplexity, Accenture and JPMorgan Chase, are already using it.
Originally reported by Seoul Economic Daily. Article compiled by the Deccan Age desk.
