The Four-Field Model: Why AI Resistance Never Looks the Same Twice

Most AI projects don’t fail because of the software. But that statement alone doesn’t help much as long as it’s unclear what they actually fail on. The honest answer is more uncomfortable than any technical explanation: they fail on resistance. And within one and the same company, that resistance can show four different faces in four different places.

That’s exactly why blanket “change management measures” so rarely work for AI adoption. A workshop designed for senior leadership misses the real problem facing employees — and vice versa. To untangle this, I’ve been working for some time with a simple framework that combines two questions.

Two Axes, Four Fields

The first axis is the hierarchy level: are we talking about a leader or an employee? The second axis is the person’s relationship to AI: are they an active user of an AI tool — or are they the object of external control, evaluation, and monitoring by an AI system?

Combine these two axes and you get four fields, each carrying its own psychological dynamic:

Field 1 – Leaders as users. Anyone who defines leadership through decision-making authority and expertise experiences an AI that delivers better analysis as helpful at first — but soon also as a threat to their own role. A typical line from these conversations: “There’s no way software can judge this better than I can with 20 years of experience.”

Field 2 – Leaders as affected parties. As soon as a leader’s own performance becomes visible through AI dashboards or metrics, that same status anxiety intensifies — only this time, the control doesn’t come from above, from their own leadership tier, but from below: from data generated within their own team and made visible upward by the system.

Field 3 – Employees as users. Here, self-doubt dominates, not status anxiety: anyone unsure of their own AI competence tends to avoid the tool rather than risk embarrassment — boasting about AI results as if they were their own, or deliberately hoarding knowledge to make themselves indispensable, are further forms of self-protection.

Field 4 – Employees as affected parties. The field with the greatest psychological charge: when AI isn’t a tool but an instance of evaluation, the result is a feeling of being at its mercy, dehumanization, and often, as a consequence, a solid loss of trust — reactance.

Is one’s freedom of action experienced as expanded by AI, or as constrained? That single question determines acceptance or resistance across all four fields.

Why This Distinction Changes Something in Practice

The real value of this model isn’t theoretical — it shows up in your next conversation with your own team. Once it’s clear which field is at play, the right question to open it up changes too. For Field 1, it helps to ask: “What would change for you as a leader if part of the analysis no longer came from you?” For Field 4, on the other hand: “Do your employees know what data is being collected, for what purpose — and who sees it?” Two completely different conversations, for two completely different problems, that from the outside often look identical: “resistance to AI.”

The perception gap most often overlooked: leaders systematically underestimate how much — and how strongly — their teams are unsettled by AI. Less than a third believe their employees are worried at all; in reality, it affects the majority. This misjudgment itself becomes an obstacle, because obvious fears go unaddressed in the change process. The result is a kind of false success, built on mere outward compliance rather than genuine buy-in, participation, and co-creative development.

Where You Stand Right Now

If you want to find out which of the four fields — perhaps in a specific part of your organization — is currently strongest, the Four-Field Diagnostic Tool is the fastest way to get there.

This Is Just the Beginning

This article kicks off a series. In the coming weeks, I’ll dive into each of the four fields individually — with concrete signal phrases, real patterns from practice, and what actually helps in each case. The next installment, on Field 4, arrives in about a week.

Don’t miss it: get an email reminder whenever a new part of the series goes live.