SUSE and NVIDIA Launch Sovereign AI Agent Architecture for Enterprise Control
October 25, 2026 · 3 min read
SUSE has introduced new capabilities designed to allow companies to adopt private artificial intelligence solutions within corporate environments while maintaining operational control, security, and governance over their systems. This initiative reinforces enterprise security for autonomous artificial intelligence agents through the integration of the new NVIDIA OpenShell reference architecture into SUSE AI Factory with NVIDIA. The announcement marks a significant step toward providing businesses with sovereign control over their artificial intelligence operations without sacrificing performance or scalability.
The newly unveiled architecture combines several advanced hardware and software components to create a robust environment for agent execution. It utilizes NVIDIA Vera Rubin for model inference and NVIDIA Vera central processing units that run OpenShell to handle agent execution. Furthermore, the system incorporates NVIDIA BlueField-4 data processing units that execute NVIDIA Sentry, which is based on NVIDIA DOCA. This specific configuration enables independent monitoring of the host system and enforces policies directly at the silicon level, ensuring that security measures are embedded deep within the infrastructure rather than relying solely on software layers.
As a launch partner for the OpenShell reference architecture, SUSE contributes to the development of an open and industrial-scale framework intended to create, isolate, monitor, and govern autonomous artificial intelligence agents with enterprise-grade security. By leveraging isolated Kubernetes environments and application programming interface gateways alongside OpenShell, the solution provides a controlled setting that regulates agent execution. The latest version, available with SUSE AI Factory with NVIDIA, incorporates zero trust model-based security for agents. This approach is specifically designed to offer immediate compatibility with next-generation silicon platforms from NVIDIA, allowing engineering teams to deploy applications and information technology departments to govern them across various infrastructure types.
Thomas Di Giacomo, the Chief Technology and Product Officer at SUSE, stated that as autonomous artificial intelligence agents and advances in next-generation silicon platforms transform corporate operations, artificial intelligence sovereignty becomes a non-negotiable requirement. He emphasized that combining enterprise Linux designed for cutting-edge hardware with active security for agents allows information technology teams to rapidly expand artificial intelligence usage. This expansion occurs while keeping proprietary data, system availability, and budgets under strict control, addressing key concerns for organizations looking to scale their artificial intelligence initiatives safely.
The infrastructure known as SUSE Linux Enterprise Server for NVIDIA enables companies to accelerate artificial intelligence workloads by providing immediate access to the latest hardware innovations, including the NVIDIA Vera Rubin infrastructure which forms the core of the OpenShell reference architecture. This consolidated and reliable enterprise Linux platform is designed to scale smoothly up to rack-scale architectures. It effectively connects the rapid evolution of silicon innovations with strict requirements for operational stability and security, ensuring that businesses can manage growth without compromising on reliability or safety standards.
In the public sector, the Finance Secretariat of the Mexican state of Guanajuato is leveraging SUSE AI Factory with NVIDIA to validate sovereign artificial intelligence capabilities with the goal of building a cognitive government for its citizens. José Luis Huerta Santos, from the Information Technology Planning and Architecture area of the secretariat, noted that using this factory offers a clear path to becoming a truly cognitive government. He highlighted that having a sovereign and open-source base allows for the rapid evaluation of artificial intelligence workloads while ensuring that citizen data, infrastructure availability, and operational costs remain under strict local control.
Additionally, SUSE and Fujitsu are collaborating to offer a hybrid quantum computing platform powered by artificial intelligence. This partnership integrates SUSE AI Factory with NVIDIA and NVIDIA CUDA-Q software on Fujitsu servers. The solution aims to move hybrid quantum intelligence from research laboratories into corporate data centers. This represents the first time that enterprise-grade hybrid quantum computing and NVIDIA AI Enterprise are offered as a unified technology stack ready for production and executed on sovereign infrastructure. Udo Wuertz, the Chief Technology Officer of FSAS Technologies EMEA at Fujitsu, stated that this collaboration provides the secure infrastructure foundation companies need as they expand their technology portfolios with new quantum computers and digital annealers.