For decades, the organisation chart has been one of the clearest representations of how a company operates, it shows hierarchy, reporting lines, functional ownership and managerial accountability, it tells us where expertise sits, who is responsible for decisions and how work is expected to move through the organisation.
But the nature of work is changing.
AI agents are moving beyond simple automation: they can analyse information, interpret objectives, make recommendations, execute tasks, interact with enterprise systems and, within defined boundaries, act on behalf of a business function. They are becoming persistent participants in commercial and operational processes.
This raises an important question:
what if AI agents and employees were represented together in the organisation chart?
Not because AI agents are human, they are not, they do not possess human judgment, responsibility, empathy or leadership; but they increasingly have something in common with organisational roles: a defined purpose, a scope of activity, access to resources, measurable performance and an impact on business outcomes.
So, the organisation chart may therefore need to evolve from a hierarchy of people into a map of accountable capabilities.
The omnichannel parallel
The evolution of AI-enabled organisations can be understood through the lens of omnichannel Retail.
A boutique, an e-commerce site, a marketplace and a clienteling service are not simply different points of sale; each has a distinct role in the customer experience, a specific operating model, its own cost structure and a measurable contribution to the company’s P&L.
A store may create discovery, trust and emotional connection.
E-commerce may provide convenience, breadth and continuous availability.
Clienteling may deepen relationships and increase loyalty.
Marketplaces may extend reach, although with different implications for margin, control and brand equity.
The purpose of omnichannel management is not to maximise every channel independently, it’s to orchestrate the channels so that the total customer and enterprise value is greater than the sum of the individual parts.
The same logic can apply to a hybrid workforce.
An AI agent, a human team and a human–AI workflow can each be understood as distinct organisational capabilities: each has a role in the value chain, a level of authority, a cost base and a contribution to business performance.
The comparison is not perfect, but it is revealing.
A channel has an owner, a proposition, a service level, a cost structure and a performance contribution. So should an AI agent. A channel may compete with another channel for investment or customer interaction – Agents may create similar tensions with human teams or with one another.
The question is therefore not whether AI agents should be treated as employees. They should not. The question is whether they should be treated as managed participants in the operating model.
From software tool to organisational capability
The distinction between a conventional software tool and an AI agent is becoming increasingly important. A traditional application generally waits for a user to initiate an action, while an agent may interpret a goal, formulate a plan, use several systems, execute activities and escalate when a decision exceeds its authority.
That does not make the agent an employee. It makes it a managed digital capability.
An agent that has a persistent business purpose, access to enterprise systems, measurable performance and the ability to influence operational or commercial outcomes should have a formal organisational identity; it needs a named business owner, a technical owner, a defined scope of activity and clear limits on what it may decide or execute, it also needs a cost profile, performance indicators, an escalation path and a process for review or retirement.
Current thinking on AI-agent governance increasingly emphasises identity, access controls, human oversight, accountability and lifecycle management: in that sense, an AI agent may deserve a place on the organisation chart without occupying a human position.
Consider a customer-growth function: alongside a Customer Strategy Director and a CRM and Clienteling team, the organisation may include a segmentation agent, a next-best-action agent and a campaign-optimisation agent.
The chart would not suggest that these agents report to the director as human colleagues would, it would make visible the capabilities supporting the function and, more importantly, the human leader accountable for their use.
Giving agents a place in the P&L
If AI agents become part of the operating model, they should also become part of the economic conversation.
Too many AI initiatives are still evaluated through broad productivity claims or isolated project-level return on investment, while a more mature approach would assess the contribution of each significant agent to revenue, margin, customer experience, risk and organisational capacity.
An agent might influence revenue, protect gross margin, reduce operating cost, shorten cycle times, improve service quality or release human capacity for higher-value work. At the same time, it generates costs through model usage, platform infrastructure, data preparation, integration, maintenance, monitoring, human supervision and compliance.
Its economic contribution can therefore be expressed as:
Net Agent Contribution, Revenue and Margin Impact, Cost Avoidance, Capacity Released, Technology Cost, Oversight Cost, Risk and Remediation Cost, etc…
The phrase “capacity released” deserves particular attention and it should not automatically be interpreted as headcount reduction.
In a luxury environment, an agent that reduces administrative work may give sales associates more time to build relationships with clients, it may allow merchandising teams to identify opportunities earlier, help buyers analyse a more complex assortment or enable senior leaders to focus on strategic decisions rather than information gathering.
The value is not necessarily labour substitution, often, it is the intelligent redeployment of human capability.
The new form of channel conflict
Omnichannel organisations have long managed tensions between stores and e-commerce, wholesale and direct-to-consumer, local and central teams, and short-term conversion and long-term brand equity.
The hybrid organisation will face similar tensions between AI agents and human teams.
A conversion agent may prioritise immediate sales while the brand team protects desirability; a customer-service agent may reduce cost while weakening a premium client experience; a pricing agent may improve sell-through while compromising price integrity.
There may also be tensions between agents: one may optimise inventory availability, another margin and a third customer experience. Each could perform successfully against its own objective while collectively producing a poor business outcome.
This is why AI deployment cannot remain purely departmental.
Organisations will need enterprise-level orchestration, shared principles, authoritative data and explicit rules for resolving competing objectives.
The future challenge will not simply be to deploy individual agents, it will be to manage a portfolio of agents as an integrated organisational system.
Human accountability remains essential
The analogy between agents, employees and channels has clear limits.
An AI agent can have a role, a cost, a capacity and a set of objectives, it cannot carry moral, legal or executive accountability in the same way as a person.
A human must remain accountable for the agent’s purpose, the data it uses, the decisions it can influence, the actions it can execute and the outcomes it produces; a human must also remain responsible for the risks it creates and for the decision to scale, modify or retire it.
The future organisation chart should therefore show two levels of ownership: the capability performed by the agent and the human leader accountable for that capability.
This is not administrative complexity for its own sake, it is the foundation of trust.
Designing the hybrid enterprise
The most effective organisations may eventually manage human and AI capabilities through a common operating framework and for every significant role or capability, leaders will need to define the outcome required, the work to be performed, the level of autonomy permitted, the systems and data required, the quality standard and the points at which human intervention is mandatory.
They will also need to understand the cost of operation, the value created, the risks involved and the conditions under which the capability should be reviewed or retired.
This creates a more useful conversation about the future of work: instead of asking whether AI will replace people, leaders can ask which activities should be automated, which decisions require human judgment, where agents can increase scale and where humans must remain in direct contact with customers. They can ask which capabilities should be centralised, which should remain close to the business and how value and cost should be allocated across the organisation.
These are not merely technology questions, they are questions of organisational design, leadership and commercial strategy.
Toward a new organisational model
The organisation of the future will not be human versus machine, it will be an orchestrated portfolio of human capabilities, AI capabilities and human–AI workflows; each will have a role in the value chain, a defined level of authority and a transparent contribution to the enterprise.
Humans will continue to provide judgment, empathy, creativity, leadership, accountability and relationship capital; AI agents will provide scale, speed, continuity, pattern recognition and execution. Leaders will be responsible for designing the interfaces between the two.
For Luxury and Premium Retail, the strategic principle could be expressed simply:
AI agents should industrialise complexity so that human teams can concentrate on value, desirability and relationships.
The most important question is therefore not how many employees an agent can replace.
It is “What combination of human and AI capabilities creates the strongest customer experience, brand equity, margin and organisational resilience?
The real promise of the hybrid organization isn’t the disappearance of human work, but its deliberate reinvention.
