A board that decides a process will run "with AI" is thereby voicing an ambition that presupposes something different than the same ambition without AI. Not more capacity in the same way of working, but work that changes shape: part of it falls to a system, part remains with a human who approves or rejects with reason, and part stays human work as it was. That division is not fixed before you look. It differs by task, by process, by company.
The operational readiness gate therefore asks a question that precedes any pronouncement: can this organization, as it is currently structured, actually let the work that the ambition presupposes take place. Not whether it is desirable. Not whether competitors are already doing it. But whether the systems, the data, the roles and the decision rights are positioned so that the work can pass through them.
In one company it is already established today who approves an AI outcome, based on which data, and what happens when the outcome is rejected. In another company that question does not yet exist, because the work has never been split into the part a system can do and the part that requires oversight. That difference rarely lies in the technology. It lies in whether the organization already knows its own work at the level of the task, rather than at the level of the function or the process.
A job title says little about readiness. A task does: is the input structured or not, is there an established rule for what counts as a good outcome, and is there someone with the mandate to reject an outcome without the process grinding to a halt. Companies that have already answered these questions per task move faster, not because they are bolder, but because they already have the evidence in hand.
The test does not ask for a plan or an intention. It asks for indications that already exist, across eight dimensions that together determine whether the ambition can land operationally: from the quality of the underlying data to the clarity of decision rights, from the existing margin for error in the process to the question of whether someone with authority stands ready to challenge an AI outcome.
A comparable standard applies per dimension. Sufficient evidence is not a positive feeling at the boardroom table, but a concrete example: a task that has already been partially taken over and for which it is known how often the human intervenes, a data flow that has already been checked for completeness, a role whose mandate is on paper rather than in the culture. Five confidence gates mark where that evidence is missing, making clear not that the ambition is wrong, but where it still rests on assumptions rather than on established work.
The question of which work in this specific company can actually be taken over by AI is answered by FTE TO AI's work scan at the level of the task, not the function.
A readiness test can conclude: not yet, evidence is missing here. That is not a verdict on the ambition, and it is not advice on what should happen to people if the work does shift. Where an outcome touches on personnel decisions, its own legal requirements apply; these are not filled in or replaced here.
What the outcome does do is point to the place where the ambition and the organization do not yet fit together. Sometimes that is a dimension that can be strengthened with relatively limited effort: a missing definition of a good outcome, a decision right that exists but has not been recorded. Sometimes it points to something more fundamental, such as data that is not in order or people who cannot yet carry the work in its new form. Where that suspicion arises, the testing continues via a test of whether your data is good enough for AI or via a test of whether your people can handle it, depending on where the evidence is thinnest.
Operational readiness does not stand apart from the other layers of the ambition. An ambition that is operationally feasible but does not affect the margin is a different ambition than what the board had in mind; for that there is a test of the margin feasibility of an AI ambition. And an ambition that is operationally and financially sustainable can still founder on what is legally permitted with the data and the decisions the work requires, which a test of the legal feasibility of an AI ambition exposes. These gates do not stand apart from one another; they test the same ambition from different angles.
Before the full ambition test is deployed, with its four layers and eight dimensions, there is a short first step: a free readiness check of eight questions, one per dimension, which gives a picture of where the organization is furthest along and where it is least so. Not a judgment on the ambition itself, but a first place to start looking.
The full ambition test, with the capability translation back to roles and decision rights, is under construction.
Vertel wat u wilt bereiken, dan kijken we samen wat daarvoor moet staan.
Answers come from this site’s knowledge base. Not tailored advice, and not a scan of your company.