Between November 2025 and February 2026, the RAG4Edu partners surveyed 147 educators about generative AI and Retrieval-Augmented Generation (RAG) in their teaching. The survey was complemented by a review of recent research on RAG in higher education.
147educators surveyed, Nov 2025 – Feb 2026
Highinterest in RAG — low confidence in building it
Key findings
- Most respondents want to use RAG in teaching but do not yet, because they do not know how it works.
- Main obstacles: lack of RAG knowledge, limited time, insufficient technical and institutional support, no pedagogical examples or templates, and GDPR concerns.
- Clear preference for modular, self-paced and blended formats, mentoring, communities of practice, and pathways that fit different entry levels.
- Demand for shared repositories of RAG templates, a European learning platform and recognition through transferable micro-credentials.
What this means for RAG4Edu
The findings shape the RAG4Edu Curriculum and Training Programme: short modules that educators can take at their own pace, group activities and mentoring in the Community of Practice, templates and worked examples, and transferable micro-credentials.
The survey was available in English, German, Spanish, Catalan, French, Portuguese, Finnish and Ukrainian.