10 Key Insights into NVIDIA and SAP's AI Agent Security Framework

At SAP Sapphire, NVIDIA and SAP announced an expanded collaboration to bring trust, security, and governance to specialized AI agents operating within enterprise systems. From finance to supply chain, these autonomous agents are critical for businesses but require robust guardrails. Here are ten essential things you need to know about this groundbreaking partnership.

1. Specialized AI Agents Are Entering Enterprise Systems

AI agents are no longer just assistants; they now handle tasks in finance, procurement, supply chain, and manufacturing. These agents operate where business decisions are made, data is accessed, and workflows run at scale. With this shift, enterprises must ensure that agents can act autonomously while maintaining security and compliance. The collaboration between NVIDIA and SAP focuses on enabling these specialized agents with the necessary boundaries and policy enforcement.

2. SAP Sapphire Announcement Highlights the Partnership

During SAP Sapphire, NVIDIA founder and CEO Jensen Huang joined SAP CEO Christian Klein's keynote via video to unveil the expanded partnership. The announcement emphasized how enterprises can now run specialized agents with enhanced security and governance controls. This marks a significant step in making agentic AI practical and trustworthy for large-scale business operations.

3. NVIDIA OpenShell Powers Secure Agent Deployments

NVIDIA OpenShell is an open source runtime designed for securely developing and deploying autonomous AI agents. SAP is embedding OpenShell into its Business AI Platform to provide isolated execution environments. It enforces policies at the filesystem and network layers, offering infrastructure-level containment. This prevents damage when agent logic fails, ensuring that agents operate within safe limits.

4. SAP Engineers Co-Design OpenShell with NVIDIA

SAP engineers are actively collaborating with NVIDIA to further develop the OpenShell open source codebase. They focus on contributions that address enterprise needs for production-grade agentic AI. This includes runtime hardening, policy modeling, enterprise identity integration, and auditing and governance hooks. The joint effort ensures that the runtime meets real-world security demands.

5. OpenShell as the Runtime Security Layer for SAP AI Agents

Within the SAP Business AI Platform, OpenShell serves as the runtime security layer for all AI agents. This includes custom agents built in Joule Studio, SAP's environment for creating and managing enterprise agents. By providing a central security foundation, OpenShell enables consistent policy enforcement across all agent activities, from finance to logistics.

6. The Trust Equation Changes with Autonomous Agents

Moving from AI assistants to autonomous agents requires a new trust framework. Agents that can touch systems of record, cross application boundaries, and operate without human review need clear boundaries, policy enforcement, and audit trails. This is exactly what the SAP and NVIDIA collaboration addresses, ensuring agents can be trusted in production workflows.

7. The Application Layer Is Critical for AI Value

NVIDIA CEO Jensen Huang describes AI as a five-layer cake: energy, chips, infrastructure, models, and applications. At the top, applications create economic value and drive productivity. SAP is a global leader in enterprise applications for finance, procurement, supply chain, and manufacturing. Its position at the core of enterprise operations makes it a key driver for adopting agentic AI.

8. Business Agents Need Role and Process Awareness

For agents to function safely, they must understand roles, processes, permissions, and data boundaries. They also need an execution environment that limits what an agent sees, what it can do, and where inference runs. The collaboration ensures that agents are context-aware and operate within defined constraints, reducing the risk of unauthorized actions.

9. NVIDIA's Perspective as a Longstanding SAP Customer

NVIDIA itself runs finance, supply chain, and logistics on SAP systems. This gives both companies a shared understanding of what enterprise-grade governance requires in practice. The real-world experience helps shape the development of OpenShell and other technologies, ensuring they meet the demands of complex business environments.

10. A Foundation for Trustworthy Agentic AI Deployment

The combination of SAP's application expertise and NVIDIA's technology provides a robust foundation for deploying trustworthy agentic AI. By addressing security, governance, and policy enforcement, enterprises can confidently adopt autonomous agents. This partnership marks a significant milestone in making AI agents safe and effective for production-scale operations.

In conclusion, the expanded collaboration between NVIDIA and SAP is set to transform how enterprises deploy specialized AI agents. With OpenShell as the security backbone, businesses can now implement autonomous agents with the necessary controls to ensure trust and compliance. As the industry moves toward agentic AI, this framework provides the guardrails needed for successful enterprise adoption.