AI in the classroom: IT:U launches school workshops on trustworthy AI

Why can an AI answer sound convincing and still be wrong? How can you spot mistakes, and when is human judgment needed? With the 2026/27 school year, IT:U, together with 506.ai and INCHSTONE, launched the school workshops of the TRUST-AI4Schools research project. At five partner schools in Upper Austria, students from primary through upper secondary school are testing a critical and responsible approach to AI. Over the project period through fall 2028, more than 60 workshops are planned.

Understand, develop, apply

“Universities have a responsibility to make scientific knowledge usable for people. With TRUST-AI4Schools, we give students the opportunity to experiment with AI themselves, question its potential and limits, and actively help shape how we use it,” says Stefanie Lindstaedt, Founding President of IT:U.

The workshops combine three building blocks that build on each other.

Understanding: Students explore how AI systems generate answers and why errors can occur.
Developing: They design age-appropriate AI applications, formulate prompts, and examine how different sources and inputs influence the results.
Applying and reflecting: Children work with examples and data (e.g., images), analyze them with AI, and critically review the outputs. This makes it tangible how AI learns from data, what its results are based on, and how reliably they can be assessed.

“To judge AI outputs, you need to understand how data, the question, and the system interact. When students collect their own data and analyze it with AI, they can investigate this connection. They learn which statements are supported by their data and where the limits of its explanatory power lie,” explains Bernd Resch, Professor of GeoSocial AI at IT:U and the project’s scientific supervisor.

“It’s not access to AI that will determine young people’s futures, but how they use it: checking, contextualizing, asking questions. You don’t learn that ‘sometime later,’ but now, in school. That’s why we provide schools with a protected environment in which every answer shows its source, without real names and in compliance with the GDPR. An AI is only as good as the sources it’s given,” emphasizes Gerhard Kürner, CEO and founder of 506.ai.

Seven guiding questions

The formats are based on the seven requirements of the Ethics Guidelines for Trustworthy AI developed by a European Commission-appointed expert group. In the workshop, these become age-appropriate questions:

Does the human keep the decision?
How reliably and safely does the system work?
How is my data protected?
How can I recognize AI-generated content, and how transparent are its foundations?
Are people being disadvantaged?
What are the consequences for society and the environment?
Who takes responsibility for errors?

A particular focus is on equal opportunity and diversity. The offerings take into account different prior knowledge, perspectives, and lived realities, and are intended to encourage especially girls and young women to actively help shape technology.

“Through our application-oriented workshops, we want to help children understand: AI is a tool, not a substitute for thinking. Human judgment remains irreplaceable. That’s why we place special emphasis on fostering independent thinking, creativity, and a responsible approach to AI,” says Sindhuja Enzinger, CEO of INCHSTONE and the project’s Diversity Lead.

Scientifically accompanied: Materials for classroom use

IT:U is studying how participants’ AI literacy develops over the course of the project. The focus is on recognizing incorrect or fabricated AI outputs, so-called hallucinations, as well as the ability to evaluate results with justification and to use AI responsibly. Participating classes will be accompanied throughout the project period. The findings will feed into the further development of the workshops, academic publications, and freely available teaching materials.

“Blind trust in AI is no more helpful than blanket rejection. Students learn to ask questions, check sources, and decide for themselves when to follow an answer, and when not to. We are studying how their AI literacy and their understanding of responsible AI use develop in the process,” concludes Heidrun Mühle, Project Lead of TRUST-AI4Schools at IT:U.

TRUST-AI4Schools runs until September 2028 and is funded through the FFG program “Talente regional 2025,” with support from the Federal Ministry for Innovation, Mobility and Infrastructure (BMIMI).

Beyond the partner schools, classes can request workshops at the Digital City Studio of IT:U and the City of Linz.