Ivory & Co.
AEA by Ivory & Co.StandardsKnowledge & ResearchImplementationCertificationIvory & Co.
Ivory & Co./Standards/Publication 15 · Governance 02 of 02

AEA by Ivory & Co. Governance System Design

Designing an enterprise that can evolve without losing its integrity.

How do you govern an enterprise that is designed to change?

Traditional governance has been organised around control.

Policies establish what is permitted. Approval processes determine what may proceed. Committees review proposed changes. Registers record decisions, and assurance examines whether requirements were followed. These mechanisms serve important purposes.

Enterprises, though, are becoming far more dynamic. They operate through people, intelligent agents, technology platforms, outside partners and distributed capabilities. Their arrangements change quickly, and a single change can affect many processes, capabilities, relationships, systems and obligations. They also need freedom to experiment and to find new ways of creating value.

Governance therefore has to hold two requirements that look contradictory: the freedom to innovate, and the obligation to protect what becomes consequential. The AEA Governance System Design establishes how they coexist. Its governing principle is: govern the consequence, not the creativity.

Governance is a system, not another architecture

Adaptive Enterprise Architecture describes an enterprise through eleven architectural dimensions. Governance is not a twelfth. It is a subject that can itself be architected using those eleven.

Directional Architecture helps establish the intent of Governance. Domain Architecture identifies the capabilities it requires. Execution Architecture describes how governance work is performed. Workforce Architecture establishes participation and authority. Trust Architecture examines when reliance on governed outcomes is warranted. The remaining architectures each add their own view. The result is an integrated operating design, where most organisations have a collection of disconnected approval procedures.

Standardise protection, not implementation

A multinational, a small professional services firm and an autonomous software operation all need governance. They do not need identical machinery.

Two enterprises may both need to establish that a consequential change was authorised. One uses a formal approval committee. The other relies on a designated decision authority supported by structured records and automated controls. The mechanisms differ, and the requirement to establish legitimate authority is the same. Governance standardises the protection and leaves the implementation to the context, which keeps it from prescribing technology, structure or workflow.

The boundary between innovation and embedment

A business investigates three possible operating models: expanding its workforce, sourcing specialist capability through partners, or using intelligent agents for substantial portions of the work. During exploration the alternatives may be incomplete, contradictory or speculative. That is no governance problem.

The design protects the distinction between innovation and embedment. The originating environment decides when exploratory thinking is developed enough to be handed forward, and Governance does not prescribe how that environment operates. Once a matter enters structured alignment, it may progress, return for further investigation, divide, converge with another proposal or be discontinued. Not every possibility needs to become governed change.

Understanding before determination

A proposal to replace an established process with an automated alternative may be technically feasible. Feasibility does not establish whether the change suits the enterprise. The business may need to know which instructions would be affected, which capabilities depend on them, who participates, what obligations and conditions of reliance would change, and what conditions actually exist today.

The design recognises a pattern of movement: Innovation, where understanding diverges and possibilities expand; Enablement, where submitted thinking is aligned, interpreted and challenged towards a determination; and Absorption, where the implications are understood against different downstream conditions. This is a pattern specific to Governance. It is no universal lifecycle and need not run in one sequence; governed work may involve nested and interacting loops.

After a change is approved

Approval and embedment are different events. An approved change may not have been implemented. An implementation may differ from what was approved. Some parts of the enterprise may adopt the new arrangement while others continue under an earlier baseline.

The design distinguishes assessment, determination, approval, implementation and actual operating consequence, and recognises that downstream contexts may legitimately keep different baselines or variants. The aim is to know which conditions apply, where differences exist and what follows from them. Immediate synchronisation is not required.

Consequences depend on actuality

An enterprise introduces a new version of an operating standard. One business unit uses the current baseline. Another retains an approved earlier variant. A third has made local changes nobody fully understands. The consequences cannot be established from the proposed design alone. Where actuality is unknown, that uncertainty stays visible, and nobody may assume every unit conforms to the intended architecture.

Traceable divergence

Change does not always require uniform adoption. A product may have several versions. A business unit may run an approved local variation. A customer implementation may carry specific conditions. When a source proposition changes, the downstream consequences are assessed, and each dependent context is treated on its merits. That is the difference between preserving architectural coherence and demanding operational uniformity.

Governance across relationships

Enterprises depend on suppliers, technology providers, contractors, partners and increasingly autonomous systems. Where a capability is partly delivered by an outside provider and partly by intelligent agents, the enterprise needs to know which obligations originate in its own architecture, which arise from contracts and which depend on the provider’s arrangements, and who holds authority for which decisions. Governance requires visibility of those relationships without becoming the owner of every actor or obligation involved.

Governance that people and machines can use

A governance system cannot depend on people remembering which document holds the rule, and a machine-readable record is not meaningful merely because software can retrieve it. Governed information must keep enough identity, context, authority, applicability and lineage to be interpreted correctly. A board needs a concise view of consequential decisions. An operating team needs instructions and approval conditions. An intelligent agent needs machine-readable requirements and explicit authority boundaries. The representations differ and the governing truth beneath them is the same.

The thirteen Governance Standards

The design is accompanied by thirteen normative Governance Standards covering the system itself, governed objects and records, identity, authority, sources, intellectual property, lifecycle, change, dependencies, releases, information, assurance and repositories. The standards establish the protections required wherever Governance is instantiated. The design explains the rationale and the relationships through which those protections are understood.

Assurance must examine effectiveness

An organisation can complete approval forms, maintain registers and hold periodic reviews while never establishing whether its protections work. A completed protocol is no proof that reliance is warranted, and the presence of an outside reviewer does not establish independence. Assurance should be proportionate to the consequence involved and should preserve relevant uncertainty.

An enterprise should be free to discover its future without losing control of what becomes its reality.

Publication record

Publication
AEA Governance System Design
Version
v1.0.1
Status
Published baseline — editorial correction. Clarifies present-state design language without changing the approved design.
Steward
Ivory & Co.
Integrity
SHA-256 85a6f89766ba7deeee3e3ddedee666db1a0e6aa3c42a93d32820465279b91a88
Related
Governance Standards · AEA Design · Architecture Set Standard
Download PDF© 2026 Ivory & Co. · Licence terms to be confirmed
·Full publication

Governance System Design v1.0.1

The complete text, in its original section structure and governed wording.

Adaptive Enterprise Architecture --- Governance System Design

v1.0.1 --- Published Baseline — Editorial Correction

Status: Published — v1.0.1. This editorial correction clarifies present-state design language without changing the approved Governance System design.

1. Purpose

This document defines the AEA Governance System as it is designed to operate.

The Governance System is a governed system that can be applied wherever governed consequence exists. It is not an AEA Architecture and is not confined to a single layer, artefact type, product, implementation or organisational context.

The Governance System uses the eleven AEA Architecture Standards to make Governance itself architecturally legible. The resulting design is expressed here as an integrated operating design. The architectural lenses explain the dimensions of that design; they do not describe a build sequence.

The Governance Standards define the normative protections through which this Design is governed and applied. The Governance System Design and Governance Standards therefore have different functions:

  • Governance System Design defines the structure, operating logic, relationships and design properties of the Governance System.
  • Governance Standards define the normative protection requirements that apply wherever Governance is instantiated.

2. Design boundary

Governance is not a twelfth AEA Architecture and is not exclusive to the canonical AEA layer. Governance is a subject that can itself be architected using AEA and then applied wherever governed consequence exists.

The design objective is therefore not identical governance machinery everywhere. It is a coherent governing system capable of preserving common protections across very different governed contexts.

Governance standardises protection, not implementation.

3. Governance System design through the eleven AEA architectural dimensions

Directional Architecture — Intent

The Directional dimension establishes the governing intent: protect the integrity, legitimacy, traceability and safe evolution of governed subjects without suppressing legitimate innovation, contextual specialisation or downstream divergence. Governance must preserve consequence-appropriate protection while keeping execution as simple as possible.

Design implications include: govern consequence rather than creativity; preserve canonical authority without assuming downstream authority; permit legitimate retained baselines and variants; and make protection durable across change.

Domain Architecture — Scope

The Domain dimension establishes the capability scope the Governance System must possess. The current thirteen Governance concerns collectively cover system governance; governed objects/records; identity/addressing; authority/decision governance; source/provenance; IP; lifecycle/status; change; dependency/blast radius/integrity; version/release/publication; information/schema/machine readability; assurance/conformance; and repository/access/retention.

The thirteen-parent set remains flat. Additional complexity or implementation detail does not justify a new parent concern unless a necessary governance phenomenon has no coherent owner.

Ecosystem Architecture — Interconnectedness

The Ecosystem dimension establishes that Governance operates across relationships among governed subjects, authorities, sources, owners, implementers, commercial interpreters, contractual parties, machines/agents and downstream derivatives. Rights and obligations may originate outside Governance; Governance consumes them as conditions.

Governance therefore needs explicit relationship, exchange, dependency and lineage visibility without assuming ownership of the actors or subjects involved.

Execution Architecture — Instructions

The Execution dimension establishes that Governance requires bounded executable mechanisms for handoff, alignment, determination, change, release, assurance, exception handling, downstream consequence assessment and re-entry. These mechanisms can be recursive, composable and specialised.

Activity is not progression. Evidence is not determination. Approval is not embedment. Assessment is not mutation. Operational stages or funnel visualisations are mechanisms/renders, not the ontology of Governance itself.

Expression Architecture — Individuality

The Expression dimension establishes that the same governing truth may require different consumable forms for Standards, products, software, SKUs, customer implementations, humans, machines, boards, certification and public material.

Renders and publications must not manufacture governing truth. Expression may vary while the governing protection and proposition lineage remain coherent.

Possibility Architecture — Optionality

The Possibility dimension establishes the deliberate boundary between exploratory possibility and governed embedment. Innovation may remain messy, contradictory and unconverged. The originating innovation context determines when thinking is sufficiently formed to be intentionally handed forward. Governance does not pull from or prescribe the innovation environment.

Once handed forward, possibilities may progress, hold, return, split, merge, redirect, converge, terminate or re-enter. Most possibilities need not become governed change.

Reality Architecture — Actuality

The Reality dimension establishes that Governance cannot determine consequence solely from intended design or current canon. It must be able to establish the actual position of governed subjects: what exists, what baseline or variant is operating, what has actually changed, what is unknown and where evidence is incomplete.

This becomes particularly important downstream. Blast radius identifies potential consequence; Reality supplies the actual state against which consequence can be established. Unknown actuality must remain explicit rather than being invented.

Intelligence Architecture — Understanding

The Intelligence dimension establishes a Governance-specific movement pattern:

Innovation --- divergent understanding → Enablement --- convergent understanding → Absorption --- divergent downstream understanding

In Innovation, possibility expands. In Enablement, submitted thinking is aligned, interpreted, challenged and progressively reduced toward consequential determination. In Absorption, an embedment or source evolution is interpreted against actual existing conditions to understand different downstream consequences.

This is a Governance System Design pattern expressed through Intelligence Architecture. It is specific to the Governance subject unless separately established as a universal Intelligence Architecture proposition. It is not a universal linear lifecycle: nested, recursive and interacting loops may operate throughout.

Intelligence owns the integration/understanding, not the subjects being understood. Governance consumes that understanding to protect consequential transitions and determinations.

Workforce Architecture — Participation

The Workforce dimension establishes that Governance must distinguish roles, participants and authority. Humans, agents and organisations may participate in governance work, but technical capability or task execution does not itself create authority.

Governance therefore needs explicit role requirements, participant assignment, delegated bounds and escalation conditions while preserving the established AEA distinction between Role, Participant, Actor and Authority.

Environment Architecture — Enablement

The Environment dimension establishes the physical and technological enablement required for Governance to operate: authoritative structured knowledge, stable identity/addressing, history/version preservation, controlled change, machine-readable relationships, monitoring/testing and reliable human/machine publication.

This use of Environment enablement is distinct from the Governance-specific Enablement movement label expressed through the Intelligence dimension. Machine semantics and renders must disambiguate them by context.

Trust Architecture — Reliance

The Trust dimension establishes that Governance exists partly to make warranted reliance possible. Reliance cannot be created by assertion, completed forms or protocol presence alone. The protections relevant to the consequence must actually be effective, their authority and currency knowable, and uncertainty visible.

Assurance therefore tests protection effectiveness, not governance theatre. Externality alone does not prove independence, and a known defect or unknown condition may qualify reliance rather than forcing false certainty.

4. Integrated Governance System design

The Governance System has the following integrated design properties:

  • Orthogonal applicability: Governance can apply to any governed subject or layer; it is not itself an AEA Architecture.
  • Common protection, contextual implementation: protective intent is inherited; implementation may specialise.
  • Consequence-based engagement: Governance protects consequential movement without colonising innovation.
  • Applicability-based inheritance: rules propagate according to applicability and lineage, not fixed hierarchy.
  • Destination-specific composition: each governed destination resolves its applicable protection set and authority.
  • Traceable divergence: source evolution triggers consequence assessment, not forced synchronisation.
  • Actuality before downstream determination: Reality informs what actually exists; uncertainty is preserved.
  • Portfolio legibility without authority collapse: cumulative and relational consequence can be seen without centralising all decisions.
  • Machine/human addressability: governing truth, rationale, lineage, status and applicability remain traversable.
  • Effective assurance: conformance is protection effectiveness proportionate to consequence, not box completion.

5. Governance Standards structure

The Governance System is governed through thirteen parent Governance Standards:

  1. GOV-01 AEA Governance System Standard
  2. GOV-02 Governance Object & Record Standard
  3. GOV-03 Identity & Addressing Standard
  4. GOV-04 Authority & Decision Governance Standard
  5. GOV-05 Source & Provenance Governance Standard
  6. GOV-06 Intellectual Property Governance Standard
  7. GOV-07 Lifecycle & Status Standard
  8. GOV-08 Change Management Standard
  9. GOV-09 Dependency, Blast Radius & Integrity Standard
  10. GOV-10 Version, Release & Publication Standard
  11. GOV-11 Information, Schema & Machine-Readability Standard
  12. GOV-12 Assurance & Conformance Standard
  13. GOV-13 Repository, Access & Retention Standard

These Standards are not the eleven architectural lenses repeated. They are the normative governance concerns that operationalise those lenses to Governance as a subject.

6. Standard → protection protocol → specialisation → implementation

The semantic relationship is:

Governance Standard → owns governing protection(s) → protection protocol expresses the applicable protection for a governed subject → profile/specialisation resolves contextual requirements → implementation mechanism realises the protection operationally

A protection protocol is therefore not a fourteenth Standard and not another architectural layer. It is the applicable governing expression of protection owned by one or more Governance Standards. Cross-standard composition is permitted where a destination requires several protections; ownership remains with the relevant Standards.

Specialisation may alter implementation detail, thresholds, evidence, timing, participants or mechanisms where permitted, but may not silently defeat inherited protective intent.

7. Design-to-Standard traceability rule

Each material normative Governance proposition should be traceable to the Design requirement(s) that justify it. Traceability does not make a Design proposition normative; it makes the relationship between design intent and governing protection legible. Conversely, an architectural observation should not be promoted into a source AEA Standard merely because it was useful in designing Governance. Universal source-Standard change requires its own architectural case.

8. Open implementation decisions

The Design does not yet lock final canonical identifier grammar, universal lifecycle vocabulary, final disposition terms, certification model, technical schema, repository technology, assurance scoring, contractual terms or a universal protection checklist. Those remain implementation, profile or governed design decisions to be resolved when their applicable context requires them.

Keep readingKnowledge is a body of meaning, relationships and authority.Standards · Body of Knowledge →
→Continue

Where to go next.