25 August 2026 · René de Baaij
Why Leadership Matters All the More in the Age of AI
By René de Baaij and Auke de Jongh
When machines take over predicting and producing, one question remains: why do people actually work? That is exactly where the task of leadership in the age of AI begins.
It happened in an ordinary team meeting. Someone demonstrated an AI tool that did in seconds what had taken the team an afternoon. People laughed, and someone made the familiar joke: soon the AI will be doing all of this anyway. More laughter. And then came a silence that lasted just a little too long.
That silence is the real subject of this article. Beneath the joke lies not a question about technology but a question about value and the right to exist: will what we can do still matter? Whoever hears only an implementation issue in that silence misses what is truly at stake.
Economists such as Ajay Agrawal, Joshua Gans and Avi Goldfarb aptly describe AI as a prediction technology: it makes prediction radically cheap and forces organizations to redesign their decision-making. Who may decide, on what basis, and what remains human work? Most organizations answer these questions instrumentally: licences, prompt training, productivity measurements. All sensible, and all insufficient.
Because while the tools are being rolled out, something happens in the undercurrent that rarely makes the agenda. People wonder whether their craft still counts. The lawyer who spent twenty years learning to draft with precision watches a model do it in seconds. The analyst who took pride in her overview sees that overview being automated. The disruption of AI is not primarily technical. It is existential.
To understand why it hits so hard, it helps to take a detour through the psychology of motivation. In decades of research, Edward Deci and Richard Ryan showed that people do not work primarily for money or status but out of three basic psychological needs: autonomy, competence and relatedness. Viktor Frankl went one layer deeper: human beings are creatures who seek meaning, and it is precisely in times of uncertainty that the question of meaning becomes urgent.
Set those three needs against the practice of implementing AI and the problem comes into view. Autonomy comes under pressure where the algorithm pre-sorts and the human is mostly left to correct. Competence wavers where skills that took years to build are suddenly replicated by a system that never tires. Relatedness thins out where colleagues increasingly work with an interface rather than with one another. Not one pillar is affected, but all three at once.
This is where the tension arises that many organizations feel without being able to name it. The figures improve, the dashboards turn green, and still something seeps away: initiative, pride, the willingness to go that extra step. An organization that introduces AI without addressing the motivational layer hollows out its own foundations, even as productivity rises.
And that brings us to leadership. Not as a job title, but as a capacity. In The Fifth Discipline, published in 1990, long before a single language model existed, Peter Senge described what organizations need in times of fundamental change: the capacity to learn faster than their environment changes. With the arrival of AI, his learning organization is more relevant than ever.
Three of Senge's disciplines deserve particular attention. The first is personal mastery: the ability to develop a personal vision and to use the tension between that vision and current reality as a creative engine. That is precisely what no model can take over. A system can predict, formulate and optimize, but it cannot want. Leaders who help people articulate what they themselves want to contribute give AI a place in that story rather than the other way round. Will you be working for the system, or will the system be working for your purpose?
The second discipline concerns mental models. Our assumptions about AI steer our behaviour more powerfully than any policy document. Those who see AI as a threat start defending and hiding. Those who see AI as a tool start experimenting and sharing. In most organizations these assumptions remain unspoken, and that is exactly why they rule. The leader Senge had in mind does not impose a vision but makes assumptions discussable, and slows down to do so if necessary.
The third is the combination of shared vision and systems thinking. Senge described the leader not as a hero but as a designer, a steward and a teacher. Nobody grasps AI alone: no director, no expert, no regulator. That turns collective learning from an appealing ideal into a hard necessity. An organization that learns faster than technology changes need not fear the technology.
What does this mean in practice for those who lead? Three movements, and none of them is a project with an end date.
The first: institutionalize the conversation about meaning. Not once at roll-out, but as a recurring rhythm, alongside the figures rather than after them. Where that conversation is missing, people fill the silence themselves, usually with their bleakest scenario.
The second: redesign decision-making together with the people who know the work. AI shifts decisions, and who gets to decide what is too important to leave to technology or to staff departments alone. Place mandate low in the organization and be transparent about which choices remain with people, and why. Those who help decide how AI is used experience its use as a choice rather than as something that happens to them. That bears directly on autonomy, the first pillar of motivation.
The third: organize learning as infrastructure, not as training. Moments for reflection in the diary, feedback loops that actually close, language that can carry tension. Treating mistakes as learning data rather than as a question of blame. This sounds soft, but it is the hardest condition there is. Without this foundation every organization falls back on old reflexes under pressure, and AI only turns up the pressure.
We are honest about our own doubts. We do not know how fast this technology will continue to develop, and nobody who claims to know deserves your full trust. Perhaps we overestimate the impact in the short term. But the underlying question does not change. Organizations ultimately exist because people want to achieve together something they cannot achieve alone. Technology can strengthen that interplay or hollow it out, and which of the two it will be is not a technical outcome but a choice of leadership.
That brings us back to the oldest question of the profession, one Senge was asking before a single chatbot existed: not what can we automate, but what do we want to create together. AI does not make that question redundant. AI makes it inescapable.
And back to that team meeting. The silence after the joke lasted a few seconds, and then the meeting moved on. That is what usually happens. But those few seconds held an invitation: to ask the question everyone felt and nobody voiced. What remains of our work if the machine can do this, and what do we want to remain? The leader who opens that conversation does something no system can do: he turns uncertainty into a shared task. That is where leadership in the age of AI begins. Not with the tool, but with the silence that follows it.
Notes
- Peter Senge, The Fifth Discipline: The Art and Practice of the Learning Organization (1990)
- Edward Deci and Richard Ryan, The "What" and "Why" of Goal Pursuits: Human Needs and the Self-Determination of Behavior, Psychological Inquiry (2000)
- Viktor Frankl, Man's Search for Meaning (1946)
- Ajay Agrawal, Joshua Gans and Avi Goldfarb, Power and Prediction: The Disruptive Economics of Artificial Intelligence (2022)
- Paul Daugherty and H. James Wilson, Human + Machine: Reimagining Work in the Age of AI (2018)