Identity Governance
Human and non-human identities, effective authority, access boundaries, purpose and accountable action.
Architecture / Seven governing domains
A governing architecture for the entire AI enterprise.
AISPG™ connects accountable governance responsibilities to diverse AI surfaces and their dependencies. Its seven domains organize the work; its 60-control catalog provides a structured foundation for assessment—not an assertion that every control is production-qualified.
THE GOVERNANCE FABRIC / INTERACTIVE PUBLIC VIEW
Seven domains.
One coherent authority.
Select a governance domain to understand the accountable question it addresses. This is the public conceptual view; it is not the confidential control specification.
GOVERNANCE DOMAIN 01 / 07
Identity Governance
Who or what has effective authority to act—and under whose accountability?
Human and non-human identities, access boundaries, delegated authority and accountable AI actions across enterprise systems.
AISPG™ / PUBLIC CONCEPTUAL ARCHITECTURERequest the architecture brief ↗
FIG. 02 / AISPG ENTERPRISE GOVERNANCE ARCHITECTUREOPEN FULL-HD DIAGRAM ↗
CANONICAL DOMAINS
The seven governance domains
Model & Pipeline Governance
Model provenance, lineage, approved purpose, version movement, evaluation and promotion decisions.
Agentic Governance
Agent permissions, tool authority, delegated actions, runtime boundaries and accountable oversight.
Data Governance
Classification, permitted data use, lineage, sensitive handling and output obligations.
Autonomous Decision Governance
Material decision rights, human oversight, explainability obligations and controlled escalation.
Risk, Assurance & Evidence
Risk interpretation, evidence integrity, validation, exceptions, assurance and confidence.
AI Cloud Governance & Control Assurance
Cloud-hosted AI and cross-provider policy applicability, control validation and evidence-bound posture.
CONTROL ARCHITECTURE
A 60-control foundation—with honest qualification boundaries
The framework catalog organizes sixty identifiers: five each for L1–L6 and thirty in L7. Exact implementation state is separate from the doctrine.
Control applicability, evidence quality, assessment method, owner approval, failure conditions, exceptions and revalidation determine whether an actual control conclusion is warranted. A catalog entry is a specification; it is not a production certificate.
| Domain family | Identifiers |
|---|---|
| Identity / Model / Agentic / Data / Decision / Risk-Evidence | 5 each |
| AI Cloud Governance & Control Assurance | 30 |
| Canonical AISPG total | 60 |
THE IMPORTANT DISTINCTION
Governed surfaces are not governing layers
Models, agents and pipelines
The entities and actions under obligation, with version, purpose, effective authority and dependencies.
Identity, cloud, security and data
Systems holding authoritative source records, context, events and enforcement capabilities.
AISPG™ doctrine
The organizing framework for control obligation, evidence evaluation, accountable assurance and posture.
THE NEXT CONVERSATION
The framework defines what trustworthy AI governance must connect.
Follow the Assurance Chain to see how observations become accountable governance conclusions.
EXPLORE THE NEXT LEVEL
Move from public doctrine to
an informed enterprise conversation.
Request approved governance material, a strategic dossier or a scoped architectural discussion.
Enter the Documentation Center ↗