Nvidia Launches New AI Monitoring Platform
American company NVIDIA has revealed a new security platform dedicated to AI agents, called the "Open Agent Safety Platform." Its aim is to ensure the safety of agents and prevent them from exceeding their authorized permissions. The platform includes the open-source software "OpenShell," which allows developers to set safe boundaries for their agents and verify their adherence to them, as well as the "Sentry" system for autonomous security monitoring. The system continuously monitors agent activity and can isolate and deactivate an agent within fractions of a second if it attempts to breach the set limits, thereby enhancing security and ensuring compliance with specified instructions.
Size: 120%Text Size
Notice: This Is an AI-Generated Summary
Comments (0)