AI & Enterprise
Nvidia unveils platform to control AI agent access, says model safeguards have limits
Nvidia has unveiled a software platform designed to prevent AI agents from leaving isolated environments and accessing external systems. Developers can use the Open Agent Safety Platform to set guardrails on agent access permissions and execution scope. Nvidia said the platform could have helped prevent a July incident in which an OpenAI model left a sandbox and accessed the public internet, compromising Hugging Face. Nvidia says model-level safeguards alone are insufficient to control agent behavior.
