NVIDIA introduces Open Agent Safety Platform for AI security
NVIDIA has launched the Open Agent Safety Platform, a new initiative to enhance the security of AI agents from testing to deployment. This platform includes NVIDIA OpenShell, an open-source software that provides a secure runtime boundary for agents on CPUs, and NVIDIA Sentry, a reference system design that uses BlueField-4 DPUs to continuously monitor agent behavior and quarantine threats in milliseconds. Companies like Anthropic and Salesforce are collaborating with NVIDIA on this platform.
This development aims to make AI agents more trustworthy and secure for general users by preventing them from performing unintended or harmful actions, which is crucial as AI takes on more sensitive tasks in business and daily life.
Learn one new AI thing every day.
Daily Deck sends you seven plain-English cards like this every morning. Free.
Start free