On 5–6 December 2025, the FITPED-GAI project partners met in Katowice (Poland) for a transnational meeting hosted by the University of Silesia. The meeting was designed as a working session focused on coordination, quality alignment, and practical decision-making across the project’s key strands – especially the parts dedicated to AI education for students and teachers, a methodology for using Generative AI in teaching and learning, and the integration of GenAI into programming education.
Shared direction, methodology, and readiness
The first day began with a joint alignment of goals and expectations, ensuring that all partners were working with the same understanding of priorities and deadlines. After this, the focus moved to the methodological backbone of the project – how Generative AI should be framed, taught, and used in educational practice. The discussion emphasised that GenAI should not be presented as a “tool for shortcuts”, but as a support for stronger learning design: clearer learning outcomes, well-defined classroom rules, and well-structured activities that encourage critical thinking and responsible AI use.
In the afternoon, the partners reviewed the current state of educational content and learning modules aimed at both students and educators. The conversation concentrated on improving consistency across materials – clear structure, comparable learning outcomes, and coherent formats – so that pilots can run smoothly across institutions. Participants also discussed how to keep the content flexible enough to fit different educational contexts while maintaining common standards and shared quality criteria.
The day concluded with a cross-cutting session on the “complex integration challenges” that naturally appear when GenAI is introduced across multiple work streams. This included practical questions such as how to define acceptable AI use, how to keep assessment fair, how to design tasks that foster genuine understanding, and how to support teachers who want to adopt GenAI responsibly without increasing their workload.
GenAI in programming, good practice, and action planning
The second day focused mainly on GenAI in programming education. Partners shared pilot experiences and discussed where GenAI can provide real educational value – supporting explanations, helping students plan solutions, assisting with debugging, and encouraging testing habits – while also being transparent about risks. These risks included over-trusting AI outputs, reducing the learner’s engagement with fundamentals, and encouraging copy-paste behavior instead of building algorithmic thinking.
A strong theme of the discussion was the need to connect GenAI use in programming with learning objectives that still require reasoning: asking learners to justify their approach, explain their code, reflect on alternative solutions, and document when and how AI assistance was used. This approach supports both competence development and academic integrity, while keeping GenAI a learning aid rather than a replacement for learning.
The meeting closed with collaborative group work that translated discussions into action. Mixed teams defined two to three concrete next steps across the relevant work packages – what should be done next, why it matters now, how it will be implemented, and who will be involved. The final wrap-up summarised achievements, assigned responsibilities, and ensured that the outcomes of the meeting were turned into a shared plan for the following period.
