Nvidia agent safety push raises questions about who governs AI autonomy

AI Business
Read full story· 6 min
Nvidia launched the Open Agent Safety Platform to help enterprises control autonomous AI agents' access and actions amid growing concerns about AI governance gaps. The platform includes OpenShell and Sentry to independently enforce limits on agent autonomy, with over 100 organizations involved in its adoption. This move addresses the need for controls beyond model-level safeguards as AI agents gain more autonomy.

More on this story


More in Cybersecurity