There is an image in the work of Dr. Gerald Hüther that I return to regularly in training sessions, because it is precise and immediately understood at the same time. A stool with three legs. Whoever sits on it sits stably, even on uneven ground, even when things shake. Whoever loses one leg has to hold on constantly, uses energy simply to avoid falling, and no longer has free hands for what actually needs to be done.

The three legs represent three forms of trust: trust in one's own abilities, trust in other people and in the network that holds one up, and trust in the meaningfulness of what is currently happening. Hüther does not describe these three resources as personality traits, nor as the result of positive thinking. They are neurobiological conditions. When they are present, the brain is able to think in a differentiated way, to learn and to make genuine decisions even under pressure. When they are absent, it shifts into protection mode. And then what I described in the first article begins to run: fight, flight or freeze.

What this has to do with AI implementations

AI implementations interfere with all three legs simultaneously. The first leg wobbles because people suddenly are no longer certain that their existing competencies are still sufficient. The second wobbles because roles shift, responsibilities become unclear and the feeling arises that one can no longer rely on familiar routines and familiar colleagues. The third loses its footing when no one explains where all of this is heading, what these changes mean and whether one's own contribution still counts in this new context.

The result is not resistance in the classical sense. It is a stool that wobbles. And people sitting on a wobbling stool no longer have spare capacity for creativity, for genuine collaboration or for courageous decisions. They are occupied with not falling over.

What leaders can do

Hüther writes that trust cannot be prescribed. It is formed through experience. In concrete terms: people need situations in which they experience that their competence matters, even when the task is new. They need colleagues and managers who listen before they explain. And they need a direction that is communicated honestly, even when not everything is yet clear. Precisely this last leg, trust in the meaningfulness of what is happening, is the one most neglected in most AI projects. Strategy presentations do not substitute for an answer to the question people are actually asking themselves: am I still part of this?

A stable stool does not come from more information. It comes from experiences that nourish all three forms of trust at the same time. That is not a soft leadership task. It is the real one.