NVIDIA Launches Open Agent Safety Platform to Secure AI Agents
NVIDIA has introduced an open software platform and reference system design to enhance AI agent security from testing to deployment.
NVIDIA today announced the launch of the NVIDIA Open Agent Safety Platform, an open software platform and reference system design aimed at strengthening AI agent security from testing to deployment. The platform consists of two key components: NVIDIA OpenShell, an open-source software providing a secure runtime boundary that traces all actions and enforces policy as agents run on NVIDIA Vera CPUs, and the NVIDIA Sentry reference system design, which continuously monitors agent behavior and can quarantine agents attempting to move outside their boundaries in milliseconds.
OpenShell can be extended to work with third-party compute platforms, including those from Arm and Intel, making it a versatile solution for various hardware configurations. The Sentry system design runs on NVIDIA BlueField-4 DPUs to ensure real-time monitoring and quick response to security threats.
Industry leaders such as Anthropic, Cisco, CrowdStrike, Dell Technologies, Figure, HPE, Hugging Face, JPMorganChase, Microsoft, Palantir, Palo Alto Networks, Perplexity, Red Hat, Salesforce, SAP, Scale AI, ServiceNow, and SpaceXAI are collaborating with NVIDIA to enhance AI safety across the full stack of infrastructure, software, models, and robotics.
The NVIDIA Open Agent Safety Platform is available through the NVIDIA developer resources page and GitHub. This initiative supports the mission of the Open Secure AI Alliance, which aims to strengthen AI agent security through open research, skills, and tools, fostering international cooperation and aligning on evaluation methods.
Source: nvidia-newsroom