Milestone for digital sovereignty: secunet, Red Hat and deepset present AI offering for classified information based on NVIDIA technologies. 10/26/2026 A consortium of secunet, Red Hat, and deepset is developing a cloud offering with AI capabilities for authorities and organizations with classified information requirements, based on technologies from NVIDIA. The partners confirmed this at NVIDIA GTC Berlin that took place from October 20-22, 2026. The new service will in the future be operated and offered by secunet cloud in its own German data centers and according to the highest standards. Additional partners are invited to provide applications, so that a strong ecosystem will gradually develop around the solution. Specifically, the future offering provides a curated, protected Runtime as a Service for AI applications that is compliant with the legal requirements for classified information (Verschlusssachen - VS). This model offers maximum flexibility and scalability for customers. The product combines the trustworthiness and sovereignty of a national cloud service provider - secunet cloud - with globally leading technology. To this end, the consortium partners are joining forces. The future offering combines the technologies and competencies of the four partners into an end-to-end approach for the secure and sovereign use of modern AI. The special requirements for the protection and processing of classified information are an integral part of the architecture. Open technologies and standardized interfaces create the technological foundation for this. secunet acts as the cloud service provider and contributes its decades of expertise in classified IT as well as its experience as the operator of the highly secure, classified-information-compliant SINA Cloud. Red Hat is the global pioneer and leading player in the field of open-source operating systems and Kubernetes architectures, providing the Red Hat AI Factory with NVIDIA platform, which integrates NVIDIA technology. deepset brings a global standard for AI agent orchestration into play with Haystack, provided sovereignly from Germany and specially designed for protected environments. Architecture combines performance and high security Technologically, the proven security architecture of secunet's SINA Cloud forms the foundation of the new solution. It is built directly on the NVIDIA BlueField-3 DPU (data processing unit). By offloading and strictly isolating security tasks at the hardware level, the solution fulfills the requirements of the Classified Information Instructions (Verschlusssachenanweisung - VSA) for handling classified data. Building on this, Red Hat AI Factory with NVIDIA ensures high-performance, containerized cloud operations and maximizes the efficiency of the NVIDIA AI infrastructure and the NVIDIA Nemotron open models deployed on it, which guarantee full transparency and adaptability. The security components of the underlying SINA Cloud are authorized by the German Federal Office for Information Security (BSI) for use with classified information. This facilitates the classified information approval of specific implementations of the solution. Digital sovereignty through openness and control Digital sovereignty is a central design principle of the new offering. Authorities and organizations should be able to retain control over which models, data, applications, and technologies they use and how they are operated. The modular structure allows components to be further developed or exchanged and the architecture to be supplemented with additional technologies and applications. Voices of the partners "The use of modern AI in conjunction with classified information requires performance, security, and sovereignty to be considered together from the very beginning. Our cloud solution for AI, which we are developing with our strong partners and will offer as part of the secunet cloud portfolio, meets this demand." Andreas Rückriegel, Vice President Cloud Services, secunet "Open technologies create freedom of choice and transparency and form an essential foundation for digital sovereignty. With Red Hat AI Factory with NVIDIA, we are bringing an open platform for the development and productive operation of AI into the new classified-information-compliant offering. Together, we combine openness and scalability with the special requirements of security-critical application areas." Christof Orth, Director Sales Federal, Local States Government and HealthCare, Red Hat GmbH "The concrete benefit of generative and agentic AI arises in applications that connect models with relevant knowledge, data, and tools. With Haystack, we are bringing an open technological foundation for the development and orchestration of such applications into the new offering and creating the prerequisites for making RAG and agentic applications usable even for use cases with classified information." Milos Rusic, CEO and co-founder, deepset Next steps and expansion of the partner ecosystem The technological foundation for the new AI cloud offering has already been laid. In the next step, the components will be integrated and their interaction for the intended deployment scenarios will be validated. In parallel, the partners are concretizing the offering built upon it. Following successful integration and validation, as well as the necessary approval and release procedures, secunet will provide the offering.

deepset unveils reference architecture for self-improving, sovereign AI agents on NVIDIA infrastructure. New reference architecture combines open models, modular agent orchestration, and governed execution on infrastructure organizations control. deepset, the company behind the open source Haystack AI orchestration framework and Haystack Enterprise Platform, unveiled a new open reference architecture developed with NVIDIA open models and software for self-improving AI agents designed to run in locally controlled and air-gapped environments. The architecture is designed for organizations that need AI systems to adapt continuously while keeping sensitive data and workloads under their control. Developed as part of an ongoing technical collaboration between the two companies, the architecture combines open models and NVIDIA technologies for model serving, optimization and governed execution with Haystack's modular agent orchestration. Cybersecurity is the first proving ground. At NVIDIA GTC Berlin, the deepset will demonstrate the architecture through a cybersecurity agent that analyzes confidential source code for vulnerabilities and uses feedback from prior runs to refine its performance over time. From "Deploy and Forget" to Continuous Improvement At the center of the architecture are two Haystack agents that form a continuous feedback loop. The first performs the domain task. In the GTC demonstration, the agent analyzes confidential source code and vulnerability information within a locally controlled environment, using open models rather than sending sensitive code to external APIs. A second Haystack agent analyzes traces and results from those runs and proposes changes to the agent harness. Each proposed change is evaluated against predefined objectives and the current version before it can be promoted, with approval, audit and rollback controls keeping people in charge of what reaches production. The architecture combines four building blocks: open models for reasoning, a modular orchestration layer that connects those models to an organization's data and tools, an optimization environment where improvements can be evaluated, and security controls governing what agents are allowed to do. Oct 23, 2026 Prev Next 1 of 43,785 Haystack provides the open source orchestration layer and agent harness, managing how models interact with context, retrieval, tools and workflows and making those components independently adaptable. running.The solution uses NVIDIA Nemotron open models served through NVIDIA NIM microservices, NVIDIA NeMo Gym as the optimization environment, and NVIDIA NemoClaw open blueprints, which run the agent inside NVIDIA OpenShell - a secure runtime that enforces policy-based controls over files, networks, credentials, and tools. Nemotron 3 is a family of high-efficiency, open multimodal models with leading accuracy to deliver fastest time to completion for long-running multi-agent systems. "Powerful open models create the foundation for sovereign AI, but realizing their full potential requires engineering the complete system around them," said Milos Rusic, CEO and co-founder of deepset. "By combining NVIDIA's open models and optimization technologies with Haystack's agent orchestration, organizations in Europe and beyond can build specialized agents that evolve with their environment while operating within governed infrastructure they control, and under human oversight." "Organizations need agents that work effectively while retaining control over their data, infrastructure and operations," said Pat Lee, vice president of strategic partnerships at NVIDIA. deepset is providing a reference architecture organizations can deploy themselves, with the controls regulated industries require." Sovereign by Architecture For governments and regulated organizations in Europe and elsewhere, sovereignty requires more than running a model locally. Organizations also need control over what data and context an agent can access, which models and tools it uses, and the infrastructure on which the complete system operates. The reference architecture is designed to preserve that control across the stack. Open models can run on locally controlled NVIDIA-accelerated infrastructure, including NVIDIA DGX Spark; Haystack keeps the agent harness modular across models, tools, and data sources; and OpenShell provides isolation and policy controls around agent execution. Haystack's modular architecture allows organizations to adapt individual parts of the system as their requirements change, reducing dependence on a single model provider or closed application stack. [To share your insights with AiThority, please write to [email protected]]