SANTA CLARA, CALIFORNIA / RankWire.AI / – NVIDIA has introduced its Open Agent Safety Platform designed to bolster security measures for autonomous AI agents. This system encompasses testing, deployment, software, computing infrastructure, and robotics components. NVIDIA announced the technology on September 28, as businesses increasingly adopt agents within complex computing environments. The platform integrates open source runtime software with autonomous hardware monitoring capabilities. NVIDIA explained that organizations can tailor the deployment of individual platform modules based on their specific security and operational needs.
OpenShell functions as the platform’s secure runtime environment, establishing enforceable boundaries for AI agents. This open source software monitors agent activities and enforces policies while agents operate on CPUs. Operators are able to set access permissions for files, networks, tools, processes, and credentials prior to agent activation. During operation, OpenShell ensures these restrictions are maintained. NVIDIA highlighted that the software supports both open and closed AI models, creating a security boundary that exists outside the model and its controlling agent.
The second key element, NVIDIA Sentry, offers autonomous monitoring via BlueField-4 data processing units. Sentry functions independently of the agent’s software environment, continuously observing its behavior. The system is capable of quarantining an agent within milliseconds if it attempts to breach its authorized boundary. It enforces security policies autonomously through dedicated hardware. NVIDIA developed Sentry on its DOCA software platform, which provides features such as agent identity verification, access control, telemetry, and inspection of agent requests and responses.
Runtime controls extend beyond the model
This platform integrates application software, secure runtime safeguards, and computing infrastructure into a layered security architecture. OpenShell isolates agents within sandboxed environments using kernel-level separation and policy enforcement. Sentry introduces an additional hardware layer beyond the agent’s control. Recent security incidents that bypassed application-level controls have prompted NVIDIA to design a system that offers organizations controls separate from both the AI agent and the guiding model.
Over 100 organizations are collaborating with technologies associated with the Open Agent Safety Platform, NVIDIA reports. Anthropic has partnered with the company to develop extra controls for managed AI agents and sandbox environments. Salesforce has integrated OpenShell with Slack, enabling teams to review agent activities, audit logs, and permission requests. The platform’s participants span sectors including financial services, enterprise software, cybersecurity, cloud infrastructure, robotics, and energy. It also supports security measures for autonomous systems operating in physical environments.
Open source software broadens deployment prospects
OpenShell is compatible with NVIDIA Vera, the company’s CPU designed specifically for agentic AI workloads. Developers are permitted to extend the open source software to third-party platforms, including systems utilizing Arm and Intel architectures. The overall platform can function across various hardware environments, with Sentry relying on BlueField-4 for its independent monitoring functions. NVIDIA states that the software for the Open Agent Safety Platform, including OpenShell and related components, is accessible through its developer resources and the GitHub platform.
Jensen Huang, NVIDIA’s founder and CEO, emphasized that AI safety and security present complex full-stack engineering challenges. The platform also contributes technological advancements to the Open Secure AI Alliance, an initiative managed by the Linux Foundation. This alliance emphasizes open research, tools, and security protocols for AI agents. NVIDIA affirmed that its platform supports governance across software, hardware, computing, and robotics layers. The release introduces new runtime and hardware controls, facilitating the deployment of autonomous AI agents across an expanding array of business systems.