The Control Room Is Not Empty
Why AI governance is already writing itself into everyday life
I have been thinking about how institutions make power feel dangerous to use.
Not reckless power. Not domination dressed up as leadership. I mean the legitimate, necessary, already-authorized power people hold inside organizations and then somehow learn not to exercise when the stakes become moral, social, or human.
Lead With Alignment is supported by subscribers like you.
This pattern is hard to miss in AI governance.
Leaders authorize systems, fund them, scale them, profit from them, and then, when harm appears, describe themselves as passengers. The explanation is familiar by now: the system was too complex, the model too opaque, the vendor too confident, the dashboard too clean, the pace of innovation too fast for ordinary accountability to keep up.
But complexity is not innocence. Lack of transparency is not an alibi.
That is the premise behind the special issue I had the privilege of guest editing for Leadership & Organization Development Journal: “The Architecture of Accountability: Algorithmic Autonomy and the Discipline of Stewardship.”
The formal guest editorial does the scholarly work of naming the architecture: the Kaplan Paradox, the control room, the manual override, and the discipline of stewardship. What I want to do here is something slightly different. I want to explain why that architecture is already touching ordinary life.
Because AI governance is not abstract.
For anyone who hears “AI governance research” and thinks it has nothing to do with everyday life — or assumes the work is too technical, too academic, or too difficult to read — I would ask you to look again.
This work is present in the résumé that never reaches a human being, the school platform no one can explain in plain language, the benefits form that rejects a person without a conversation, and the medical score, credit score, risk score, productivity score, or customer profile that gradually narrows what options are available.
It is present when a chatbot becomes the front door to a service that used to be reachable through a person. It is present when a workplace system turns human contribution into surveillance residue. It is present when recommendation engines shape what children see, what communities argue about, what doctors notice, what employers count, and what institutions come to treat as true.
AI governance is not only about technology. It is about who gets seen, who gets sorted, who gets believed, and who is allowed a future.
These systems are not like the weather. They do not simply arrive. They are built through procurement choices, budget approvals, design defaults, vendor contracts, leadership incentives, missing escalation paths, and all the small permissions that teach an institution what it is willing to tolerate.
And because they are built, they can be changed.
A nine-month journey
This special issue was the result of a nine-month journey that began at the International Leadership Association Virtual AI Summit in 2025, following an editorial panel discussion I facilitated.
Special issues are often described through their finished products: the articles, abstracts, tables of contents, and polished arguments. What disappears is the labor of formation—the correspondence, review cycles, revisions, coordination, encouragement, deadlines, editorial judgment, and intellectual care that make a collection cohere.
This issue arose because many people took the work seriously.
I am grateful to Stefanie Johnson for her early vision in appointing me as Guest Editor and for giving me the intellectual freedom to shape the issue's vision. I am also grateful to Rickard Enstroem, whose steady stewardship helped support the issue as it moved through the journal process.
Suzanne Joy Clark deserves special recognition. Her assistance in providing peer reviews was indispensable to the quality of the final collection. This kind of editorial labor is often invisible from the outside, but anyone who has shepherded a scholarly project knows how much depends on the people who keep the work moving without making themselves the center of it.
I also want to thank Lauren Malone and Emma Ferguson for their patient assistance in navigating the administrative and technical hurdles of a two-volume production.
Of course, the deepest intellectual contribution belongs to the authors.
The authors in this issue did the hard work of thinking carefully about leadership at a moment when AI is too often framed through spectacle, fear, efficiency, or inevitability. They refused to treat leadership as a personality trait or AI as a neutral tool, in contrast to much of the current work on the topic. Instead, they examined the conditions under which technical systems become sources of organizational power and asked what forms of accountability, judgment, stewardship, and resistance are required as decisions are increasingly mediated by machines.
The anonymous peer reviewers also deserve recognition. Their labor is easy to miss because it is not attached to a byline. But rigorous review is one of the ways scholarship disciplines itself. In this issue, the reviewers served as auditors of logic, pressing the arguments toward greater precision, accountability, and conceptual force.
That matters because the central governance question is no longer simply whether a tool works.
It is what kind of institution the tool helps produce.
The problem this issue takes up
For years, one of the most convenient defenses in technology leadership has been: “I didn’t know the algorithm was doing that.”
That defense is collapsing.
As AI systems reach further into hiring, education, healthcare, finance, public services, and everyday life, leaders cannot claim authority over the benefits while remaining ignorant of the harms. They cannot celebrate scale when it produces growth and then disown scale when it produces damage.
This is what I call the Kaplan Paradox: a condition in which a leader’s systemic liability expands as algorithmic reach grows, even while operational transparency and perceived managerial control shrink.
In plainer language: the farther the system reaches, the more leaders are tempted to claim they cannot see what it is doing.
But if an organization can procure the system, integrate it into workflows, train people to use it, measure productivity with it, and extract value from it, then the system is not outside leadership's control. It is part of leadership’s operating environment.
That is why this special issue is not a collection of leadership tips. It is a diagnostic audit for a new era of accountability.
It asks researchers, leadership educators, and practitioners to stop treating leaders as adjacent to AI systems and start holding them as part of the system itself.
Why ordinary governance language is not enough
We have all seen the corporate response to an AI scandal: the ethics board, the empathy workshop, the principles statement, the fairness slide, the task force, the public promise to listen.
These can matter. I am not against them.
But they can also calm the room without changing the rules. They can reassure the public, comfort the institution, and give leaders language for concern while leaving the authorization logic intact.
And authorization logic is where governance actually lives.
It lives in the incentives, thresholds, escalation rights, redress pathways, procurement criteria, audit trails, and decision rules that determine whether a system can be paused, challenged, repaired, or stopped. If those do not change, governance has not happened. The symptoms have been narrated. They have not been treated.
This is where academic research can serve a public function, not by making the issue more complicated, but by refusing to let complexity become a hiding place.
The value of scholarship is not that it floats above ordinary life. At its best, scholarship gives us the language to see what ordinary life has already been teaching us.
The question of power
One of the easiest ways systems preserve themselves is by making people tired, isolated, confused, and unsure whether action is possible. In organizations, complexity can perform that same function. It can make capable people behave as if they have no authority.
But the problem is rarely the total absence of power. More often, it is the selective use of power.
Leaders override systems all the time. They override timelines, user research, internal dissent, and risk concerns when the launch matters enough. They know how to intervene when a release date is threatened, when a brand is at risk, when investors are watching, when growth targets are slipping, or when executive preference demands movement.
So the question is not whether leaders have override power. They do.
The question is why that power appears so readily for market urgency, brand protection, and executive preference—and so hesitantly for public harm, worker dignity, gendered risk, or democratic accountability.
That is one of the tensions this issue asks us to confront: not the absence of power, but the selective use of it.
Why this matters now
Research often struggles with its public reputation. People hear “special issue” or “scholarly article” and assume the work is remote from ordinary life. Too dense. Too slow. Too hard to read. Too concerned with other scholars.
Sometimes that criticism is earned. But the answer is not to make research thinner. The answer is to make its stakes more legible.
The articles in this issue matter because they help us see what ordinary institutional language often hides. They help us ask better questions about authority, design, delegation, accountability, and repair. They give leadership scholars and practitioners a way to examine not only what AI systems do, but what organizations become when they rely on them.
That is not a niche concern. It is the work of public life now.
Every institution adopting AI is making choices about evidence, trust, discretion, voice, and consequence. Every automated workflow carries assumptions about whose judgment matters, whose suffering counts, whose exception is worth noticing, and whose future can be narrowed without explanation.
The issue is not whether AI will enter everyday life. It already has.
The issue is whether the systems shaping that life will remain hidden behind complexity, or whether we will build the governance capacity to see them, question them, and change them.
An invitation into the issue
The control room is not just a place of authority. It is a site of professional responsibility.
It is where decisions are authorized before they become defaults. It is where harm is either interrupted or rationalized. It is where leaders decide whether governance will be a living practice or a public-facing vocabulary.
The formal guest editorial develops this argument in full, along with the architecture that holds the two-volume issue together. Over the coming months, I will be featuring the articles in the issue and the hard-earned work of the authors behind them.
My hope is that this collection helps more people see AI governance not as an abstract technical concern, but as a practical question of leadership: who holds power, how it is used, when it is withheld, and what becomes possible when people stop pretending they are passengers in systems they helped build.
The control room is not empty. And the systems shaping our lives are not weather.
Hi, I’m Christine. 👋
In a world awash in buzzwords and borrowed courage, Dative.works is a pocket of resistance—where data pros, skeptics, and stray idealists come to build something less disposable.
If you’d rather ask better questions than echo easy answers, pull up a chair. The work starts here.