Scope

Domain 4 covers the design and implementation of AI operating models — the organizational structures, roles, processes, and governance mechanisms that enable an enterprise to develop, deploy, and operate AI systems at scale. It is built around the AEOM™ — the Autonomous Enterprise Operating Model — which provides the structural framework for AI operating model design.

Operating model design is the organizational layer of AI Transformation. Strategy defines where the organization is going. The operating model defines how the organization is structured to get there. Without an operating model that is designed for AI, even the best AI strategy will fail to execute.

What This Domain Covers

• The AEOM™ operating model architecture and its five structural components
• AI Center of Excellence (AI CoE) design — models, structures, and governance
• Role architecture for AI Transformation — new roles, evolved roles, and eliminated roles
• Process redesign for agentic workflows — how business processes change when AI agents are introduced
• Operating model transition planning — how to move from the current operating model to the target model
• Federated vs. centralized AI operating models — the tradeoffs and decision criteria
• The CAIO role — responsibilities, authority, and organizational positioning

Core Concepts

C4.1 — Why Operating Models Must Change

Most organizations attempting AI Transformation are trying to deploy AI systems within operating models that were designed for human execution. This creates a fundamental mismatch: the processes, roles, decision rights, and governance mechanisms of a human-centric operating model are not designed to accommodate autonomous AI systems.

The consequences of this mismatch are predictable: AI systems are deployed but not adopted, because the processes they are designed to support have not been redesigned to leverage them. AI agents are built but not trusted, because the governance mechanisms required to oversee them have not been established. AI investments are made but not measured, because the value measurement frameworks required to capture their impact have not been designed.

Operating model redesign is not optional for AI Transformation. It is the organizational prerequisite for everything else.

C4.2 — The AEOM™ Five-Component Architecture

The Autonomous Enterprise Operating Model organizes AI operating model design around five structural components:

• Component 1 — AI Leadership Structure: The executive roles, reporting relationships, and decision-making authorities that govern AI Transformation at the enterprise level. Includes the CAIO role, the AI Transformation Council, and the relationship between AI leadership and business unit leadership.
• Component 2 — AI Center of Excellence: The organizational unit responsible for developing, deploying, and governing AI capabilities across the enterprise. The AI CoE is the operational hub of the AI operating model — it provides the shared services, standards, and expertise that business units need to leverage AI effectively.
• Component 3 — Business Unit AI Integration: The structures, roles, and processes through which individual business units develop and deploy AI capabilities within their domains. Includes the AI Business Partner role, the business unit AI roadmap process, and the governance mechanisms that connect business unit AI initiatives to the enterprise AI program.
• Component 4 — AI Process Architecture: The redesigned business processes that incorporate AI agents and agentic workflows. Includes process mapping for human-AI collaboration, exception handling protocols, and the oversight mechanisms that ensure human accountability for AI-executed processes.
• Component 5 — AI Talent Architecture: The roles, skills, and career paths that the AI operating model requires. Includes both technical roles (AI engineers, ML engineers, data scientists) and business roles (AI Transformation Managers, AI Business Partners, AI Governance Leads).

C4.3 — AI Center of Excellence Models

The AI Center of Excellence is the organizational unit that provides the shared AI capabilities, standards, and expertise that the enterprise needs to transform at scale. There are three primary AI CoE models, each with distinct tradeoffs:

• Centralized CoE: All AI capability development and deployment is managed by a central team. Business units consume AI services but do not develop them independently. Advantages: consistency, quality control, efficient use of scarce AI talent. Disadvantages: slow response to business unit needs, risk of disconnect between AI capability and business context.
• Federated CoE: AI capability development is distributed across business units, with a central team providing standards, governance, and shared infrastructure. Business units have their own AI teams that develop capabilities for their specific domains. Advantages: speed, business context, ownership. Disadvantages: inconsistency, duplication, governance complexity.
• Hub-and-Spoke CoE: A central hub provides shared infrastructure, governance, and standards. Business unit spokes develop domain-specific capabilities within the framework established by the hub. This is the most common model for large enterprises — it balances the consistency of the centralized model with the speed and context of the federated model.

The choice of CoE model should be driven by the organization's size, AI maturity level, risk tolerance, and the degree of AI capability variation across business units. There is no universally correct model — the right model is the one that fits the organization's specific context.

C4.4 — The CAIO Role

The Chief AI Officer is the executive responsible for AI strategy, governance, and transformation at the enterprise level. The CAIO role is relatively new — most organizations created it between 2023 and 2025 — and its scope, authority, and organizational positioning vary significantly across organizations.

The AEBOK™ defines the CAIO role as having five core responsibilities:

• Strategy ownership: The CAIO owns the AI Transformation strategy and is accountable for its execution. This includes the AEF™ pillar roadmap, the AEMM™ maturity advancement targets, and the investment thesis.
• Governance leadership: The CAIO chairs the AI Governance Council and is accountable for the effectiveness of the AI governance framework. This includes AI policy, risk management, compliance, and oversight mechanisms.
• Capability development: The CAIO is responsible for building the AI capabilities — technical and organizational — that the transformation program requires. This includes the AI CoE, the AI talent architecture, and the AI platform infrastructure.
• Business alignment: The CAIO is responsible for ensuring that AI investments are aligned with business priorities and that AI capabilities are adopted by the business units that need them. This requires strong relationships with business unit leaders and a deep understanding of business context.
• External positioning: The CAIO represents the organization's AI capabilities and governance standards to external stakeholders — regulators, investors, partners, and the public. As AI governance becomes a regulatory and reputational issue, this responsibility is growing in importance.

C4.5 — Process Redesign for Agentic Workflows

When AI agents are introduced into business processes, the processes themselves must be redesigned — not just the technology layer. Process redesign for agentic workflows addresses three questions:

• What does the agent do? Which steps in the process are executed by the AI agent? What inputs does it receive? What outputs does it produce? What tools does it use? What decisions does it make?
• What does the human do? Which steps in the process remain with human professionals? What oversight responsibilities do humans have for agent-executed steps? What exception handling responsibilities do humans have when the agent encounters situations outside its defined scope?
• How is accountability maintained? Who is accountable for the outcomes of agent-executed processes? How are agent actions logged and auditable? How are errors detected and corrected? How are governance violations escalated?

C4.6 — Operating Model Transition Planning

Transitioning from a human-centric operating model to an AI-integrated operating model is a multi-year program that requires careful sequencing. The transition cannot happen all at once — it must be staged to manage risk, build capability, and maintain operational continuity.

Operating model transition planning follows three phases:

• Phase 1 — Foundation: Establish the AI leadership structure, design the AI CoE, and define the AI talent architecture. This phase creates the organizational infrastructure for AI Transformation without disrupting existing operations.
• Phase 2 — Integration: Redesign business processes to incorporate AI agents and agentic workflows. Deploy the AI CoE's shared capabilities to business units. Establish the governance mechanisms for AI-integrated processes. This phase is where the operating model begins to change in ways that are visible to the workforce.
• Phase 3 — Optimization: Continuously improve the AI-integrated operating model based on performance data. Expand the scope of agentic workflows. Develop the advanced human-AI collaboration capabilities that characterize the Autonomous Enterprise. This phase is ongoing — it does not have a defined end point.

Framework Alignment

AEOM™ — Primary Framework

The AEOM™ is the primary framework for D4. All operating model design work in this domain uses the five-component architecture as its organizing structure.

HWEM™ — Supporting Framework

The HWEM™ workforce evolution model is closely related to D4 — the operating model transition directly drives the workforce transformation that D6 addresses. Practitioners designing operating models must anticipate the workforce implications of each design decision.

AEGF™ — Supporting Framework

The AI governance framework (covered in D5) is a component of the operating model. Operating model design must include governance design — the two cannot be developed independently.

Practitioner Tools

Tool D4.1 — AEOM™ Current State Assessment

A structured assessment of the organization's current operating model across the five AEOM™ components. Identifies gaps between the current model and the target model required for AI Transformation.

Tool D4.2 — AI CoE Design Canvas

A structured template for designing the AI Center of Excellence. Covers CoE model selection (centralized, federated, hub-and-spoke), organizational structure, staffing model, service catalog, governance mechanisms, and funding model.

Tool D4.3 — Role Architecture Map

A visual tool for mapping the role changes required by the AI operating model. Identifies new roles to be created, existing roles to be evolved, and roles that will be eliminated or significantly reduced. Includes job description templates for the most common new AI roles.

Tool D4.4 — Process Redesign Workshop

A facilitated workshop format for redesigning business processes to incorporate AI agents. Produces a process map that clearly delineates agent responsibilities, human responsibilities, oversight mechanisms, and accountability structures.

Tool D4.5 — Operating Model Transition Roadmap

A structured template for planning the transition from the current operating model to the target model. Organizes transition initiatives across the three phases (Foundation, Integration, Optimization) with milestones, dependencies, and resource requirements.

Competency Indicators

Level 1 — Awareness

Can describe the AEOM™ five components and explain why operating model redesign is necessary for AI Transformation. Understands the three AI CoE models and their tradeoffs. Can explain the CAIO role and its five core responsibilities.

Level 2 — Practitioner

Can conduct an AEOM™ current state assessment. Can facilitate a process redesign workshop. Can produce an AI CoE design using the D4 tools. Can develop an operating model transition roadmap for a specific organizational context.

Level 3 — Architect

Can design the complete operating model architecture for an enterprise AI Transformation program. Can integrate operating model design with governance design, workforce planning, and technology architecture. Can lead the operating model transition for a complex, multi-business-unit organization.

Level 4 — Leader

Can own the operating model transformation at the enterprise level. Can make the organizational design decisions required to build an Autonomous Enterprise operating model. Can manage the political and cultural dimensions of operating model change at the executive level.

Certification Mapping

ATP™

D4 constitutes approximately 15% of the ATP™ exam. Candidates are tested on their ability to describe the AEOM™ architecture, select the appropriate AI CoE model for a given organizational context, and design a basic operating model transition plan.

EAA™

D4 constitutes approximately 35% of the EAA™ exam. The EAA™ capstone requires candidates to produce a complete AI operating model design — including CoE structure, role architecture, process redesign, and transition roadmap — for a real or hypothetical organization.

Case Study: ManufactureCo

Composite case study. All names and identifying details are anonymized.

Context

ManufactureCo is a global industrial manufacturer with $25 billion in revenue, 80,000 employees, and operations in 22 countries. The company had been deploying AI use cases for three years with mixed results — some impressive successes in predictive maintenance and quality control, but consistent failure to scale beyond individual plant deployments.

The Problem

An AEOM™ assessment revealed the root cause: ManufactureCo had a centralized AI CoE that was building AI capabilities, but no mechanism for transferring those capabilities to the 140 manufacturing plants that needed to use them. The CoE was producing excellent AI models; the plants were not adopting them because they had no AI-capable staff, no redesigned processes, and no governance mechanisms for AI-assisted operations.

The Intervention

The operating model was redesigned using the hub-and-spoke model. The central CoE became the hub — providing shared infrastructure, governance standards, and model development capability. Each of the 22 country operations received a dedicated AI Business Partner role — a practitioner who could translate CoE capabilities into plant-level deployments and redesign plant processes to incorporate AI agents.

The process redesign work was the most time-consuming component. Each plant's predictive maintenance process had to be redesigned to clearly delineate what the AI agent monitored, what it flagged, what the maintenance technician decided, and how accountability was maintained for maintenance outcomes.

The Outcome

Within 18 months of the operating model redesign, 87 of the 140 plants had deployed the predictive maintenance AI system in production. The previous three years of centralized CoE work had achieved production deployment in 12 plants.

Key Lesson

AI capability without operating model change does not produce AI transformation. The CoE can build the best AI systems in the industry — but if the operating model does not provide the structures, roles, and processes required to deploy and adopt those systems, the capability will sit unused.

Monday Morning Actions

For Practitioners — This Week

• Map your organization's current AI operating model against the five AEOM™ components. Which components are well-developed? Which are missing or underdeveloped?
• Identify one business process in your organization that has been targeted for AI augmentation. Map the current process and identify where the AI agent would operate, what the human would do, and how accountability would be maintained.
• Research the AI CoE model your organization uses (or plans to use). Is it the right model for your organization's size, maturity, and context? What would you change?

For Managers — This Month

• Conduct an AEOM™ current state assessment for your business unit or function. Identify the top three operating model gaps that are limiting AI adoption.
• Define the AI roles that your team needs but does not currently have. What would it take to create those roles — hire, reskill, or restructure?
• Facilitate a process redesign workshop for one AI-targeted process in your domain. Produce a process map that clearly delineates agent and human responsibilities.

For Executives — This Quarter

• Evaluate whether your organization has the right AI leadership structure. Is the CAIO role clearly defined and adequately positioned? Does the AI Transformation Council have the right membership and authority?
• Assess your AI CoE model. Is it producing the adoption results you need? If not, is the problem the CoE model, the CoE capability, or the business unit operating model?
• Commission an operating model transition roadmap for your AI Transformation program. Define the target operating model and the sequence of changes required to get there.