I have always been obsessed with organisational design. I think organisational design eats culture . Culture is too slow. Form always follows function. And something larg is happening in the world of organisational design. Every organisation is quietly growing a second nervous system: a synthetic layer of software that does not simply wait to be used, but increasingly senses, signals, recommends and acts. Think about what a nervous system does. It senses change. It moves signals. It coordinates parts of the body that never speak to each other directly. It triggers responses before conscious thought catches up. You do not see it on a normal anatomy sketch of the body at rest, but nothing moves properly without it. That is what synthetic labour is becoming inside the company.
Synthetic labour
Synthetic labour is labour-like capability delivered through software. It is not human, not employee, not contractor, not robot in the physical sense. But it can still perform work that used to require human time, attention, coordination and judgement. It can search, draft, monitor, compare, escalate, recommend and act. That changes the organisation.
From waiting to doing
This is the shift leaders need to understand. A spreadsheet waits. A CRM waits. A dashboard waits. Even many chatbots still wait. They respond when prompted. They are tools. An agent is different. An agent can be given a goal. It can plan steps. It can use tools. It can check data. It can monitor exceptions. It can recommend actions. In some cases, it can act. Once software can act, it enters the world of authority, accountability and trust. Who gave it the goal? What data can it use? Which systems can it touch? What decisions can it make? Who reviews the output? Who owns the mistake? That is not an IT question. That is an organisational design question.
The synthetic layer will not respect departments
A nervous system does not run along the lines of the organisation chart, and neither will synthetic labour. Agents will not sit neatly inside boxes labelled sales, finance, HR, operations, legal or customer service. They will follow the work:
- A customer-service agent may summarise customer history, recommend refunds, escalate sensitive issues and trigger follow-up actions. Is that customer service, legal, risk, brand, data or IT?
- A procurement agent may compare suppliers, check sustainability claims, flag geopolitical exposure and recommend contract language. Is that procurement, legal, finance, sustainability or risk?
- A sales agent may identify prospects, draft outreach, check CRM history, suggest pricing and monitor buyer intent. Is that sales, marketing, commercial operations, data governance or AI risk?
Agents move through flows of information, decisions, promises, exceptions and relationships. They expose something the organisation chart always hid: companies are not really collections of departments. They are systems of work. The old chart will still show formal authority. It will not show synthetic activity. That is the problem. If the chart does not show the synthetic layer, leaders will not know who owns it, who governs it, who improves it, who audits it or who is accountable when it creates consequences.
From headcount to capability
The deeper shift is from headcount planning to capability planning. The old question was: how many people do we need in this department? The new question is: what mix of human judgement, synthetic labour, automation, data, workflow design, customer contact and governance do we need to create this outcome? That is a different management logic. It moves the organisation away from departments as fixed containers and toward work as a living system. Some work should be deleted. Some should be automated. Some should be augmented. Some should be protected because it is trust-heavy, ethically sensitive or deeply human. Some should be redesigned because the workflow itself is broken.
More intelligence per unit
The shallow version of AI transformation is cost reduction. Fewer people. Faster output. Lower overhead. That is where lazy leaders will go first. The deeper opportunity is more intelligence per unit of organisation. Better sensing. Better response. Better personalisation. Better learning. Better decision support. Better risk monitoring. Better internal memory. Better orchestration.
The organisation chart needs a new dimension
An anatomy diagram that shows only the skeleton and organs, but not the nervous system, is not a real map of the body. It is a map of the body at rest. Most organisation charts are exactly that. They show hierarchy, not flow. They show reporting lines, not decision rights. They show roles, not work. They show departments, not the real movement of information, trust, risk and accountability. Future organisation charts will need another dimension. They will need to show agents, workflows, data ownership, decision rights, escalation paths, human review points and accountability. They will need to show where synthetic labour touches customers, money, contracts, employees, risk, reputation and strategy. Where every agent has an owner. Someone responsible for purpose, scope, data boundaries, action boundaries, performance, risk, review and retirement. Because the machine is not accountable. People and organisations are.
The synthetic layer is coming.
In many organisations, it is already here in the shadows. The real risk is not that leaders adopt it too early. The real risk is that they fail to see it, govern it and redesign the organisation around it.
