Control the import, verification, approval, storage, promotion, and offline build of AI software and model artifacts.
Learning outcomes
Curriculum
The chapters are individually addressable documentation pages. You can link directly to a concept from another program, architecture decision, or implementation guide.
Understand controlled artifact lifecycles and the assets and boundaries they cross.
Treat every imported package, container, model, dataset, and configuration as a controlled lifecycle object.
Inventory the data, models, prompts, artifacts, identities, tools, interfaces, infrastructure, and evidence that require protection.
Map where trust changes and where data, artifacts, requests, identities, tools, and evidence cross a boundary.
Establish verifiable identity and an approved route into authoritative internal sources.
Identify software, models, data, prompts, policy, and configuration with verifiable origin and lineage.
Design quarantine, verification, review, approval, transfer, and promotion stages with separation of duties.
Define authoritative internal sources for containers, packages, models, metadata, and dependency resolution.
Build, update, roll back, and recover without uncontrolled external dependencies.
Make disconnected builds repeatable and design controlled updates, rollback, rebuild, backup, and recovery.
Version complete AI behavior, assess change impact, approve promotion, support rollback, and retire superseded assets.
Record context, options, control consequences, evidence, dependencies, limitations, and reversal triggers for material decisions.
Define control objectives and collect evidence at each promotion boundary.
Translate sovereignty goals into testable statements of what must be allowed, prevented, approved, observed, retained, and recovered.
Design evidence alongside controls so important decisions, releases, configurations, and operating events can be verified.
Produce a supply-chain design and validate its completeness.
Design artifact quarantine, approval, promotion, rollback, and recovery evidence.
Design exercise; executable implementation pending
This public curriculum is vendor-neutral and does not authorize imports into a production environment. Product-specific commands, supported versions, signing systems, malware scanners, and transfer mechanisms must be declared and tested in separately versioned adapters.