Four minutes. That was the gap between my arrival and the closing of the examination hall door. The incident occurred during Morocco’s post-Eid examination period, when transportation disruptions are common as many workers return from their hometowns. The taxi I had taken that morning was struck by a motorbike on the way to campus. I found another vehicle, arrived as fast as I could, and was turned away without discussion. The rule was clear. Nine o’clock meant nine o’clock.
Later that day, I sat two other examinations. Both were multiple choice. I finished each in roughly ten minutes and was then required to remain seated, by regulation, for a substantial portion of the remaining time. From my chair, I watched a number of students photograph the question papers and consult AI systems on their phones. The pattern was difficult to miss. The institution had applied its rules with mechanical certainty where I was concerned. It appeared considerably less certain where academic integrity was concerned.
The instinctive interpretation is administrative failure. The more accurate interpretation is something stranger: the institution was behaving exactly as designed. The rules enforced were precisely the rules least connected to the university’s actual purpose. That is not a paradox. It is a pattern. And once you begin looking for it, it appears everywhere.
The objective that dare not speak its name
Consider the decision facing the administrator at that door. Turning away a student who is four minutes late is uncomfortable. It is also defensible. The rule exists, it was applied uniformly, and there is no record of exception. Allowing that student in creates a different problem entirely. Why this student and not the one who arrived seven minutes late? Who made the judgment call? On what basis? The moment discretion enters the picture, accountability follows. The administrator who exercises judgment becomes personally exposed in a way that the one who enforces the rule does not.
This is the operating logic of modern institutions, and it is seldom stated plainly. The institution is not asking: how do we best achieve our mission? It is asking: how do we ensure that no decision we make can be held against us? Those questions look similar from a distance. They produce entirely different organizations. One optimizes for outcomes. The other optimizes for defensibility. The tragedy is that defensibility, pursued consistently enough, destroys the outcomes it was meant to protect.
Institutions increasingly prefer a bad decision that can be defended to a good decision that requires explanation.
None of this implies that discretion is inherently superior to rules. Rules protect institutions from favoritism, inconsistency, and abuse. Universities cannot function if every policy becomes negotiable. The challenge is not choosing between rules and judgment, but preserving enough judgment to ensure that rules continue serving the mission they were created to support.
The disappearance of judgment from institutional life is therefore not a bureaucratic accident. It is a rational adaptation to incentive structures that punish being wrong far more severely than they reward being right. When accountability systems are built to identify failure rather than recognize achievement, the safest strategy is always to minimize the surface area of judgment. Replace decisions with rules. Replace rules with checklists. Replace checklists with automated enforcement. At each step, the institution becomes harder to criticize and less capable of achieving anything worth criticizing.
What artificial intelligence has made undeniable
For decades, this dynamic was visible but containable. AI has made it structural.
The examination rests on a single assumption: that the words on the page came from the mind of the student in the chair. Modern educational systems were built on top of that assumption. Every assessment rubric, every invigilation protocol, every integrity policy follows from the belief that answer production and genuine understanding are roughly the same thing. A student who could write a coherent analysis of a strategic problem was, by reasonable inference, a student who understood strategic thinking.
A language model sitting on a phone in a pocket radically weakens that inference. Answer production is no longer reliable evidence of independent understanding in the way educational systems long assumed. The examination is now measuring something different from what it claims to measure, and the gap between the claim and the reality is growing wider every semester. The students using AI in that examination hall were not the problem. They were the proof. The system had been measuring the production of answers all along. AI can now produce answers. The system has no idea what to do next, so it enforces the door-closing rule more precisely.
This is where the examination story and the broader argument converge. The institution cannot respond meaningfully to the AI problem because a meaningful response would require judgment: about what understanding actually looks like, about which assessments still measure it, about how to redesign an evaluation system that has been obsolete since the first smartphone appeared in a pocket. Judgment of that kind is slow, contested, and impossible to defend if it goes wrong. So the institution enforces what it can enforce and leaves the harder question untouched.
The legitimacy problem
This pattern is not unique to education, nor is it unique to Morocco. It appears across public administrations, universities, regulators, and large organizations throughout the region and beyond. It describes governments that produce regulations rather than outcomes. Hospitals that document compliance rather than improve care. Banks that model risk rather than understand it. International organizations that issue statements rather than act on them. In each case, the institution has substituted a defensible process for a difficult judgment, and in each case the gap between the institution’s formal purpose and its actual behavior widens quietly until something forces it into view.
What is at stake in each of these cases is legitimacy. Institutions derive their authority from a claim: that they exist to serve a purpose larger than their own survival, and that they are meaningfully engaged in doing so. Procedural rigidity, applied consistently enough, hollows out that claim. An institution can enforce its rules with perfect fairness and still certify graduates who developed nothing, regulate industries that changed without it, or administer programs that serve their own continuation rather than their stated beneficiaries.
The form of rigor and the substance of rigor are not the same thing. Institutions that confuse the two do not become more rigorous. They become more fragile. Every interaction in which the gap is visible, a student turned away for four minutes, a hall full of students consulting AI uninterrupted, is a small withdrawal from a legitimacy account that cannot be replenished by enforcing the next rule more carefully.
The cost of safety
There is a final irony worth naming. The risk aversion that drives this behavior is not irrational from the perspective of the individual administrator. The rules protect them. The procedures protect them. The decision that requires judgment is the decision that can come back to haunt them. In a world where accountability systems are faster and more punishing than ever, the safest career is the one spent enforcing the existing framework rather than questioning it.
The problem is that an institution made entirely of safe individuals becomes collectively dangerous. It loses the capacity to exercise the judgment its mission requires, precisely because each person within it has learned that judgment is a liability. The risk has not been eliminated. It has been redistributed, from the administrator who might have made the wrong call to the student who needed a right one, to the graduate whose diploma does not mean what it says, to the society that built systems of trust on the assumption that institutions were actually doing what they claimed.
A system that cannot distinguish between a student who is four minutes late and a student who is not thinking at all has not become rigorous.
It has become blind, and learned to call that blindness fairness.
