Key takeaways

1. Integration succeeds when educators understand the curriculum they are working with.
   Keynote speaker Vesna Lušicky (University of Vienna) made the case that curricula are multi-level systems — international, national, institutional, classroom, individual — shaped by values and stakeholders that are often invisible. CLARIN integration works best when it identifies the right level to act on and acknowledges what it cannot change.
   
   
2. The main barriers are structural, not technical.
   Respondents pointed to limited institutional support, lack of time, and the absence of ready-to-use materials more often than to problems with the tools themselves.
   
   
3. Teachers want ready-to-use materials.
   The clearest signals from the survey were demand for ready-to-use datasets suited to teaching, reusable slides for local adaptation, sample lesson plans, train-the-trainer sessions, and student-oriented tutorials for self-learning. The CLARIN Learning Hub is the central place where these resources are being gathered.
   
   
4. Applied use cases matter more than tool overviews. Future training should focus on connecting tools and resources to real research questions, including data analysis workflows, AI and LLM literacy, sentiment analysis, multilingual alignment, and guidance on licensing, ethics, and GDPR. Tools like the Virtual Language Observatory and the Language Resource Switchboard were highlighted as practical entry points.
   
   
5. Collaboration beyond the classroom is essential.
   Teachers alone are not the main entry point. CLARIN consortia were encouraged to work more closely with university libraries, research data management offices, and digital humanities centres, and to build a data stewardship network with expertise in language data.
   
   
6. Students are a route into the curriculum.
   Theses, data citations, and repository deposits were highlighted as practical ways to bring CLARIN into student work without redesigning whole courses.