Definitions and terminology
Use sovereignty, privacy, residency, control, portability, and assurance precisely.
Public technical foundation
Sovereign AI Body of Knowledge
A controlled, versioned foundation for designing AI systems whose boundaries, dependencies, behavior, and evidence can be understood and verified.
The public layer is designed to be useful on its own. Structured programs, production-grade labs, answer keys, maintained compatibility matrices, and customer-specific patterns remain part of private delivery.
Public scope
Technical knowledge that establishes a common control language.
Reference architecture
Separate workload, control, evidence, supply-chain, and operational responsibilities.
Trust boundaries
Map actors, data flows, import paths, runtime zones, egress, tools, and approval points.
Threat models
Reason about model, retrieval, artifact, identity, prompt, tool, and operational abuse paths.
Control checklists
Turn architectural intent into testable technical and operational control objectives.
Evidence patterns
Record what was approved, deployed, evaluated, changed, and observed.
SAI-100 · Version 1.1.0
Sovereign AI Foundations
Establish the vocabulary, deployment boundaries, shared responsibilities, and control objectives needed to reason about sovereign AI systems.
PublicContent reviewedfoundationLast content review: 2026-08-03
Learning outcomes
- Define sovereignty in terms of enforceable technical control
- Map data, model, infrastructure, operational, and evidence boundaries
- Distinguish private deployment from a demonstrably controlled AI system
- Create an initial sovereignty control objective map
- Reading time
- 29 min read
- Assessment
- Public knowledge checks and a reviewed foundation architecture worksheet
Foundation structure
01
Define the system
Scope purpose, actors, assets, data, models, dependencies, and environment.
02
Draw the boundaries
Show where trust changes and where information or control can cross.
03
Set control objectives
State what must be prevented, allowed, observed, approved, and retained.
04
Design evidence
Identify how each material decision and control can be verified.
05
Record the backlog
Turn gaps into prioritized technical and operational work.
Core principle
Sovereignty is a set of enforceable controls—not a hosting label.
A system becomes more sovereign as the organization can make, enforce, verify, and change material decisions across the dimensions below.
Data
Location, access, lineage, retention, and allowed movement
Models
Provenance, licensing, packaging, evaluation, approval, and retirement
Infrastructure
Runtime boundary, dependencies, identity, networking, storage, and hardware
Operations
Updates, observability, reliability, incidents, capacity, and recovery
Evidence
Decisions, configurations, hashes, evaluations, audit events, and verification
Reusable starter artifacts
Download the worksheets used in the foundation workshops.
These are editable starting points, not completed evidence or production approval records.
Shared module registry
Versioned building blocks for programs and future role paths.
Every module declares visibility, content maturity, and verification separately. An outline in the registry does not mean a complete lab or assessed program is available.
SAI-100PublicContent reviewed
v1.1.0Sovereign AI Foundations
Establish the vocabulary, deployment boundaries, shared responsibilities, and control objectives needed to reason about sovereign AI systems.
foundation29 min readContent review 2026-08-03
Open curriculum documentationSAI-110PublicContent reviewed
v1.0.0Trust Boundaries and Threat Modelling
Model data flows, actors, assets, attack surfaces, trust zones, and risk scenarios for private AI workloads.
intermediate27 min readContent review 2026-08-03
Open curriculum documentationSAI-120PublicContent reviewed
v1.0.0Sovereign AI Reference Architecture
Design vendor-neutral architecture layers, deployment patterns, assurance points, and decision records.
intermediate36 min readContent review 2026-08-03
Open curriculum documentationSAI-200PublicContent reviewed
v1.1.0Internal Artifact Supply Chains and Offline Builds
Control the import, verification, approval, storage, promotion, and offline build of AI software and model artifacts.
advanced33 min readContent review 2026-08-11
Open curriculum documentationSAI-210PublicDraft
v1.0.0Model Selection and Lifecycle Management
Evaluate model fit, licensing, provenance, packaging, approval, updates, and retirement inside a controlled lifecycle.
intermediate23 min readContent review 2026-08-03
Open curriculum documentationSAI-220PublicContent reviewed
v1.1.0Model Serving, Routing, and Hardware
Select and operate model runtimes, routing patterns, hardware profiles, capacity controls, and reliability targets.
advanced31 min readContent review 2026-08-11
Open curriculum documentationSAI-230PublicPrototype
v1.0.0Private RAG and Permission-Aware Knowledge
Build grounded retrieval with controlled ingestion, citations, access enforcement, evaluation, lineage, and safe no-answer behavior.
intermediate28 min readContent review 2026-08-03
Open curriculum documentationSAI-240PublicDraft
v1.0.0Identity, AI Gateways, and Egress Control
Propagate identity and enforce inspection, DLP, routing, endpoint, and response policies at controlled AI boundaries.
advanced31 min readContent review 2026-08-03
Open curriculum documentationSAI-250PublicPrototype
v1.2.0Governance, Assurance, and Evidence
Turn policies and control objectives into verifiable runtime, release, decision, and audit evidence.
intermediate41 min readContent review 2026-08-09
Open curriculum documentationSAI-260PublicDraft
v1.0.0Observability, Reliability, and FinOps
Measure workload health, behavior, evaluation quality, capacity, cost, incidents, and operational evidence.
intermediate22 min readContent review 2026-08-03
Open curriculum documentationSAI-270PublicDraft
v1.0.0Kubernetes and OpenShift Operations
Operate restricted AI workloads with controlled networking, storage, GPU access, security contexts, updates, backup, and recovery.
advanced28 min readContent review 2026-08-03
Open curriculum documentationSAI-280PublicOutline
v0.2.0Governed Agents and Controlled Tool Execution
Constrain agent tools, identity, memory, approvals, execution, and signed action evidence.
advanced32 min readContent review 2026-08-03
Open curriculum documentationPublishing boundary
Open knowledge and maintained enterprise delivery serve different needs.
Public and reusable
- Definitions and terminology
- Reference architecture overview
- Selected threat and trust-boundary examples
- Control checklists and selected technical guidance
- Introductory lab direction and open starter-kit links
Private enterprise delivery
- Complete program curricula and production-grade labs
- Lab answer keys and assessment environments
- Capstones, rubrics, and instructor support
- Maintained compatibility matrices
- Customer-specific implementation patterns
Apply the knowledge
Turn the public foundation into a team capability plan.
Use the Body of Knowledge independently, or discuss a private cohort and customer-hosted lab aligned to your roles, architecture boundary, and implementation backlog.
No public checkout, account, or sensitive architecture upload required