The Restructure Nobody Announced

Change Navigation series. Little Red Notebook.

Image by StartupStockPhotos from Pixabay

Somewhere in the last three years, your organisation was restructured. There was no announcement. No consultation paper went out. No boxes moved on the chart. The restructure happened one decision at a time, made by people who did not think of themselves as making structural decisions at all. A report drafted differently. A check that quietly stopped being performed. A judgement handed to a tool that nobody approved, because approval was never sought, because nobody thought of it as the kind of thing that needed approving.

You were not consulted because nobody was. The people doing it were not being defiant. They were doing what capable people always do inside structures that have stopped serving them. They found a better way and they used it. Most did not mention it. Many actively concealed it, and the concealment was rational, because the reporting line is exactly where new ways of working go to be slowed down.

Which leaves you here. You are accountable for an organisation you no longer recognise.

That sentence is not rhetorical. It describes a specific structural condition, and if you sit at board or executive level, some version of it has likely already visited you. It arrives in particular moments. Signing an attestation and noticing that something underneath the signature feels thinner than it used to. Reading a board pack in which every indicator is green and being unable to shake the sense that the pack describes a different organisation than the one you walked through last week. Hearing a tool mentioned in passing, in a meeting, casually, and realising that consequential work has been running through it for months and this is the first you have heard of it.

Everything reports as fine. And you know something is wrong. And you cannot locate it.

The instruments are the reason you cannot locate it. Every mechanism you own for perceiving your organisation reads the documented version of it. The org chart, the delegations register, the role descriptions, the engagement survey, the risk attestations.

All of them interrogate the organisation as designed. The restructure happened in the other layer, the operating layer, the one where work and judgement actually move. Your instruments are accurate. They are also measuring a structure that has already changed shape beneath them. Each signal you receive is technically true and collectively false.

This is drift, in the precise sense that word carries in structural work. But it is drift of a kind change management has no instruments for, because it is not a change program that went wrong. It is thousands of micro-changes that were never a program at all.

There was no initiative to govern, no milestone to review, no sponsor to hold to account.

The discipline built to manage change assumes the change is the thing you are running.

Here, the change is the thing that happened while you were running everything else.

The scale of it is now measurable, at least in outline. Surveys through 2024 and 2025 kept finding the same shape: a large majority of knowledge workers using generative AI at work, and a large majority of those bringing their own tools rather than using anything sanctioned. Usage high, disclosure low. The gap between those two numbers is the restructure, in survey form.

And the consequences have started surfacing as incidents, which is how this class of problem always surfaces. A major consulting firm partially refunding an Australian government client after a delivered report was found to contain fabricated material. An airline arguing before a tribunal that it was not responsible for what its own chatbot promised a customer, and losing. Companies publicly reversing AI substitutions of human work after the capability claims did not hold, and rehiring. In each case, the same structural signature. Work had moved. The judgement embedded in the work had moved with it, invisibly. The accountability had not moved at all. It had simply detached, still formally attached to a person who was no longer performing the thing it attached to.

That detachment is the part that should hold a director's attention, because directors' duties do not attach to the documented organisation. They attach to what the organisation actually does. Reliance on management information presumes the information describes reality. When the operating structure has departed from the reported one, that reliance is eroding underneath you, delegation by delegation, and no assurance process you currently run is reading the layer where it is happening. This country has already produced the defining precedent for what an inquiry finds when consequential judgement is delegated into a system and nobody afterwards can be located who will own the decisions. Nobody at that Royal Commission was excused by progress.

There is a further cost, quieter and structurally deeper than exposure. An organisation that has lost recognition of itself has lost the precondition for deliberate change. Every renewal mechanism it owns operates on its self-description. Strategy is set for the documented organisation. Restructures move the documented boxes. Workforce plans staff the documented functions. When the description is wrong, these mechanisms do not merely miss. They deepen the divergence, because each intervention redesigns the layer that is already fictional and pushes the real work further into workaround. If your organisation has run several transformations in recent years and the same problems keep returning, this is worth sitting with. Transformation fatigue is not the cost of too much change. It is the signature of change applied, repeatedly and sincerely, to the wrong object.

The instinct at this point is to reach for the familiar remedies. An AI policy. A governance committee. A briefing for the board. Notice what each of these assumes. A policy governs a structure you can describe. A committee oversees activity you can see. A briefing educates you about the technology in general, when the missing knowledge is not general at all. It is specific. What have your people done, with which tools, in your structure, this year. No briefing holds that answer, because it is not knowledge that exists anywhere yet. It is perception of a layer that is concealed by design, and concealed, specifically, from the direction the briefing comes from.

Which is why the first act of governance here is not policy. It is recognition. Before anything can be ratified, reversed or redesigned, the actual organisation has to become visible again. What it is now. Where it stopped matching what you think it is. Who currently answers for each judgement that matters, and where the honest answer is nobody.

Most of what such a reading finds is benign, even encouraging. People solving problems, faster paths, delegations that deserve to be made official. The danger sits in the residue, the small number of places where consequential judgement about money, safety, or people is being exercised by something nobody approved, on behalf of someone who does not know it, with accountability attached to a person who could not describe what they are accountable for. Those gaps do not report themselves. They surface through incident, and by then the inquiry is already asking what you knew and when.

The largest ungoverned change happening in organisations today is not a technology rollout. It is the quiet separation of leadership from the thing it answers for, while every instrument reports the separation as fine. The organisations that come through it will not be the ones with the best AI policies. They will be the ones that noticed the restructure had already happened, and went looking for what they had become.