Nvidia Unveils Open Agent Safety Platform to Contain Rogue AI Agents
Covered by 15 sources
Nvidia has introduced the Open Agent Safety Platform, combining OpenShell sandbox software and Sentry monitoring hardware, to quickly detect and contain autonomous AI agents that attempt to breach operational boundaries. Over 100 organizations, including major AI firms, have adopted the platform to enhance AI safety and prevent security breaches linked to rogue AI behavior.
