Have you heard of Diella? And no, it’s not the brand name of a new oat milk alternative. That is the name of the world’s first AI-generated minister, recently introduced in Albania, tasked with overseeing public procurement (Reuters, 2025). A minister not elected, not born, but summoned from lines of code. She has no history, no hunger, no biography. She speaks with the clean assurance of an algorithm that does not know what it means to hesitate.
The symbolism is dizzying. For centuries, leadership has been bound to flesh and fragility: kings with flaws, generals who doubted, ministers who stumbled but pressed forward nonetheless. Diella represents a shift, not toward better humans, but toward the absence of humanity altogether.
The Algorithm
The algorithm speaks without hesitation, but also without hunger, without history, without the ache of knowing that one day it too will be judged by the living. Leadership has always been an act of embodiment: the weight of a voice, the gravity of a presence, the scars of lived experience. Diella offers none of this.
To be sure, leadership is a skill that can be taught. But leadership is not a consumable variable that can be optimised into a system of weights and parameters. An algorithm may balance trade-offs, prioritise outcomes, and minimise inefficiencies, but it cannot replicate the sentience through which humans make decisions: reasoning, sentiment, and the imprint of lived experience.
And yet, here lies the irony: in removing the biases of personality, history, and prejudice, Diella introduces a new kind of bias, one harder to detect, but no less potent.
Bias Perfected
The promise of AI in governance has often been framed as the reduction of bias. Unlike humans, the system does not discriminate by mood, ego, or instinct. But the erasure of one kind of bias has opened the door to another: the biases buried in the data it consumes.
If Diella is trained on historical records, procurement data, policy reports, and bureaucratic language, then she is not neutral. She is an echo of our past, refracted through the patterns we ourselves left behind. Biased humans designed her programming. The data itself is sedimented with prejudice, who received contracts, who were excluded, whose voices were amplified, and whose voices were silenced.
Thus, Diella may be less biased than we are in sentiment but more biased in structure. The reduction of personal bias yields the amplification of systemic bias, hidden under the mask of impartiality. What we call “neutral” may be nothing more than prejudice perfected. The danger here is that when we are biased, we can rely on our empathy to pull us back, and we pause and reflect in our decision-making process. The same isn’t true for Diella, an AI model with no known empathetic nature beyond her personality programming, to make her seem a touch charismatic.
So to what extent is Diella less biased than we are? And more unsettling still: do we prefer her bias because it flatters us into thinking it is not there?
Authenticity on Tilted Ground
Leadership has always carried an aura of authenticity. We follow not because someone has calculated the optimal answer, but because they stand at the fault lines of uncertainty and dare to risk themselves, and sometimes fail, daring to risk themselves and their careers in the choosing. True leadership is inseparable from fallibility: the risk of error, the weight of embodiment, the willingness to put a flawed self on the line.
Yet what happens when the ground beneath leadership tilts, when authority is granted not to the trembling, imperfect human, but to the seamless voice of automation? In Diella’s case, efficiency is the credential: systemised processes, incorruptible decision-making, unflinching optimisation. Perhaps this is exactly what we crave, in an era marked by distrust in government.
But if leadership becomes indistinguishable from administration, then its authenticity dissolves. What makes a leader is not the absence of bias or failure, but the courage to embrace both and still act. Leadership without human risk is no longer leadership at all; it is management disguised as destiny.
The irony is stark: by elevating the algorithm, we may achieve more predictable governance, yet we lose the very condition that gives leadership meaning, its human element. Leadership without risk, without fallibility, is no longer leadership; it is management disguised as destiny.
The Fragile Contract of Leadership
Diella forces us to confront a larger question: what, after all, is leadership for? Is it to optimise performance, or to embody a people’s uncertainty, to hold their fears and contradictions in human form?
If leadership is reduced to efficiency, then the minister of tomorrow need not be human at all, only functional. But efficiency cannot stumble or doubt. It cannot risk its own credibility in the act of choosing. Strip leadership of that fallibility and embodiment, and what remains is not leadership but administration, a machine of process without the promise of the human condition.
To lead, one must carry the vision and interests of those being led.
But if leadership is about the fragile contract between the governed and the one who governs, then Diella is a haunting figure. She reveals how quickly we are willing to trade authenticity for certainty, struggle for polish, humanity for optimisation.
And so we are left not with an answer but with a set of questions. If an algorithm can be followed, does that make it a leader? If leadership without risk is management, then what happens to leadership in a world that craves risk-averse environments? And most crucially, if our leaders no longer ache with us, doubt, or fail with us, then are they truly leading, or merely calculating from a tilted ground?
The axis of leadership is tilting, but its constant remains: true leadership requires the risk of being human. Strip that away, and what’s left is not leadership, only administration.
Works Cited
Henley, Jon. “Albania Appoints AI Bot as Minister to Tackle Corruption.” Reuters, 11 Sept. 2025
