A qualitative study published this year by Corvinus University researcher Mátyás Hartyándi examined how an entire organizational unit takes up AI, using psychodramatic methods to surface the hidden attitudes and questions of trust that surround generative AI.
The study looked at an entire division of a Hungarian services company to see how employees take up generative artificial intelligence (ChatGPT, DALL-E, Gamma, and so on). What sets it apart is that it didn’t use questionnaires or conventional interviews but psychodramatic, projective methods: participants used objects, spatial placement, and role reversals to portray how they see AI. That made room for not only stated opinions but also hidden attitudes, fears, and inner representations to come to the surface. The study identified three main patterns:
- Experiencing AI as double — at once a tool and a threat.
- Disappointment between the performance promised and the benefit actually experienced.
- Emotional and physical distance — that is, participants didn’t fold AI into their daily work; they only “experimented” with it.
The study shows why an AI rollout can stall even when, on paper, everyone is “open” to it.
What Does This Mean for Education?
1. “Cognitive openness” on its own is worth nothing.
The most important lesson from the corporate study: someone can be open in their head — and at the same time completely resistant underneath. You can see the same in education:
- the teacher says: “I understand that AI matters”
- but inside feels: “I can’t control it,” “what happens to my role?”, “I don’t want to get it wrong”
2. Without a clear pedagogical purpose, AI “stays at a distance.”
In the study, the participants all placed AI physically far away from themselves. It’s a powerful metaphor: “If I don’t know exactly what it’s good for, I won’t touch it.”
The same holds in school:
- if the teacher doesn’t know exactly what to use it for,
- if the student sees it only as a “trick,”
- then AI never becomes part of the learning process, and it confirms the worst fears.
3. Adopting AI isn’t a technology project but an identity project.
According to the study, the main tension didn’t come from features but from roles:
- “compared to what am I now supposed to do differently?”
- “does it lower my competence?”
- “or does it actually raise it?”
In education this effect is even stronger. The change in the teacher’s role is the greatest emotional obstacle. If we don’t handle it, then no matter how much AI there is, pedagogical practice won’t change.
4. The letdown after the hype is just as present in schools.
The study shows that people expect a lot and are quickly disappointed.
For teachers, for instance, it can look like this:
- they try ChatGPT → “it doesn’t know my field”
- a hallucination → “you can’t trust it”
- a bad prompt → “more time than it’s worth”
This is NOT the teacher’s fault — there is simply no professional context built around the AI.
5. Integration happens only when there is a shared language, shared practice, shared control.
One of the study’s strongest messages: if everyone tries on their own, the organization never reaches real adoption.
In education it looks like this:
- students → private AI use
- teachers → individual experimentation
- leaders → an institutional vision, but no routine
That’s why a shared platform and a common set of concepts are the key — otherwise:
- there’s no quality,
- no safety,
- no pedagogical integration,
- no progress.