From the Red Hat edition of October 29, 2026
secunet, Red Hat and deepset present AI offering for classified information based on NVIDIA technologies
secunet Security Networks AG (with Red Hat, deepset, NVIDIA) · Sovereign government AI cloud for classified information · Germany (international comparator for SLED)
- Publisher
- secunet
- Original publication
- secunet press release, Oct. 26, 2026
- Source retrieved
- 2026-10-30
- Event date
- 2026-10-22
What happened
Vendor claim: a consortium of secunet, Red Hat and deepset, using NVIDIA technology, confirmed at NVIDIA GTC Berlin (Oct. 20 to 22, 2026) that it is developing a classified-information-compliant AI Runtime as a Service for German authorities. secunet cloud would run it in its own German data centers. It builds on the SINA Cloud architecture on NVIDIA BlueField-3 DPUs, Red Hat AI Factory with NVIDIA for containerized AI operations, Haystack for agent orchestration and Nemotron open models. secunet says the SINA Cloud security components are BSI-authorized for classified information. The new offering itself still needs integration, validation, approval and release before secunet will provide it.
Why it matters
International comparator for US state and local agencies with criminal justice, tax, health or other restricted data that want AI under government-controlled, in-country or in-state operation. It shows Red Hat AI positioned as the platform layer inside a regulated sovereign service, with accreditation inherited from an existing authorized cloud. US equivalents would map to CJIS, IRS 1075 or StateRAMP rather than German VS rules.
Evidence and measured results
Read the full secunet release (dated 10/26/2026). It is a forward-looking consortium announcement: no customers, pricing, availability date or performance data. Approval for the new offering is pending, so the BSI authorization claim covers existing SINA Cloud components, not this service.
Limitations and uncertainty
Vendor consortium announcement; not yet available; non-US regulatory context; no outcomes or customers reported.
Put this evidence to work
Lighthouse Advisory interpretation, grounded in this source. Enriched 2026-10-30; this does not change the original publication date. Each role follows the same checklist so takeaways compare across sources.
Sales
Role takeaway
- Customer problem
- Agencies holding restricted data cannot use public AI services and lack an accredited, government-controlled AI runtime.
- Stakeholders
- State CIO or shared-services director
- CISO
- Criminal justice or tax agency data owners
- Procurement office
- Discovery questions
- Which data classes block your AI use today (CJIS, FTI, health)?
- Do you have a state-operated or consortium cloud that already holds authorizations?
- Would agencies consume AI as a managed runtime from a trusted operator?
- Value hypothesis
- Building an AI runtime on an already-authorized government cloud could shorten the path to approved AI use for restricted data.
- Potential engagement
- Sovereign AI runtime architecture and accreditation-mapping workshop with a state shared-services provider.
- Unsupported claims
- That the offering is available or approved
- That any US accreditation applies
- Any performance or cost outcomes
Pre-sales engineering
Role takeaway
- Fit
- Fits shared-service providers hosting regulated workloads who want an open Kubernetes-based AI platform under their own operation.
- Architecture and integration
- DPU-isolated security foundation, Red Hat AI Factory with NVIDIA for AI runtime, Haystack for RAG/agents, open models such as Nemotron.
- Prerequisites
- Accredited or accreditable hosting environment
- GPU and DPU infrastructure
- Data classification and control mapping
- Constraints
- Not yet integrated or validated
- German VS rules do not map directly to US frameworks
- Vendor-only description of the design
- Security
- Hardware-level isolation of security functions plus in-country operation. Each implementation still needs approval.
- Proof of value
- Map each architecture layer to the target US control framework
- Pilot a non-restricted workload on the same stack before restricted data
- Document which controls inherit from the existing authorization
Delivery
Role takeaway
- Work
- Plan a sovereign AI runtime program with accreditation gates, operator responsibilities and application onboarding.
- Dependencies
- Accreditation authority timelines
- Hardware supply
- Partner application readiness
- Ownership
- Shared-services operator, with agency data owners approving use per data class.
- Skills and adoption
- Operator staff need Kubernetes AI platform, DPU and accreditation documentation skills.
- Governance checkpoints
- Authorization decision before restricted data
- Per-application onboarding review
- Periodic reassessment of controls
- Acceptance criteria
- Proposed: documented control inheritance for each layer
- Proposed: authorization granted before restricted-data use
- Proposed: onboarding process published for partner applications
- Risks
- Accreditation delays
- Vendor lock-in despite modular claims
- Uncertain demand from agencies
Implementation considerations
Lighthouse Advisory interpretation across the operating dimensions a public-sector buyer must settle before this evidence becomes a design. Each note answers the question under its heading for this specific source.
Architecture and integration
What must connect, and where does the AI sit in the workflow?
Layered design: hardware-isolated security on DPUs, then Kubernetes-based AI operations (Red Hat AI Factory with NVIDIA), then agent and RAG orchestration (Haystack), with modular, exchangeable components.
Governance
Who approves, reviews and stays accountable for outcomes?
Accreditation is approached by building on an already-authorized platform. Approval of specific implementations is still required.
Security and privacy
What data, permissions and controls need testing?
Security functions are offloaded and isolated at the DPU hardware level, and operation stays within national data centers under a national provider.
Accessibility and workforce
Who is affected, and what skills or accommodations follow?
Not addressed in the source.
Procurement
What should contracts, pricing and exit terms secure?
Delivered as a managed service from a national provider, so agencies would buy a service, not a platform build. Partners are invited to add applications to the ecosystem.
Operating model
Which teams own the service once it runs?
A government-trusted operator runs shared AI runtime for many authorities, similar to a state shared-services or consortium cloud model.
Publication history
- 2026-10-29Red Hat · Issue 018 resources
Stable resource ID: secunet-redhat-deepset-classified-ai-cloud-germany-20261026