Research Spotlight: Preparing Future Teachers for Generative AI

This tension is at the heart of the article “Exploring pre-service teachers’ attitudes and experiences with generative AI: a mixed methods study in Norwegian teacher education.” The study looks closely at Norwegian student teachers — future professionals who will decide how AI belongs in classrooms.

We spoke with one of the authors, Professor Siv M. Gamlem of Volda University College, to learn more about generative AI in teacher education.

Generative AI is now part of the context in which teachers learn, plan, write, assess, and give feedback. The article matters because it treats teacher education as the place where professional judgement about AI must be formed, not as an afterthought after tools reach classrooms.

The Norwegian context adds a useful perspective. The article notes that educational institutions in Norway are largely left to decide how to approach GenAI, which makes student teachers’ confidence, caution, and experience especially important for future policy and practice.

Professor Gamlem reflected on the broader significance of the study: “The most important implication of this research is that teacher education must take an active and deliberate role in shaping how future teachers exercise professional judgement with AI. This cannot be left to individual experimentation or informal use; it requires structured opportunities within programmes to critically evaluate when, how, and why generative AI supports learning, assessment, and feedback.”

The study’s contribution lies in showing both promise and caution. It avoids the simple story that future teachers are either enthusiastic adopters or resistant skeptics.

Awareness was widespread, expertise was not. In the survey of 209 Norwegian student teachers, 78.5 percent reported having used an AI application, yet no participant identified as an expert in AI technology. This suggests a gap between access and deep competence.

Usefulness was strongest for planning and content discovery. Respondents most often saw GenAI as useful for saving time when finding teaching material or planning lesson content, while fewer saw it as useful for reviewing homework or monitoring students in the classroom.

Experience shaped nuance. Interview participants with regular GenAI experience could describe specific uses, such as summarising difficult readings, asking follow-up questions, and drafting lesson templates, while novice users tended to express broader uncertainty and preference for human judgement.

Professor Gamlem reflected on what surprised the research team most: “We were somewhat surprised by the gap between widespread use and limited depth of understanding. Many student teachers are already using generative AI, but without a strong foundation for critically evaluating its role in teaching and learning. For teacher educators, this highlights an urgent need to move beyond access and focus on developing informed, reflective, and pedagogically grounded use of AI.”

The authors used an explanatory sequential mixed methods design. They first collected survey data from 209 student teachers across all levels of a 5-year primary and lower secondary teacher education programme at a university college in Norway, then followed up with interviews with 11 of the same participants.

The quantitative component used survey items on familiarity with AI, GenAI knowledge, beliefs about positive and negative aspects, and confidence in conventional digital technologies, then applied item response theory and regression analysis. The qualitative component used semi-structured interviews and inductive thematic analysis to compare experienced and novice users, with the research team manually checking automated transcripts and removing identifiable information before analysis.

After the initial inductive thematic analysis had been carried out by the two interviewing researchers, the anonymised interview transcripts were further analysed using Evidano.

The article states that “The research team compared the results from Evidano with those obtained from the initial manual analysis of the interviews for triangulation and validation purposes.” This wording is important because it places Evidano in a supporting role: it helped the team compare and validate patterns, while the authors retained responsibility for the analysis, interpretation, and scholarly judgement.

That use also reflects a broader issue raised by the article: responsible AI in education depends on critical human oversight. Professor Gamlem also noted how AI tools can support careful research workflows when used responsibly: “Evidano supported our research process by offering an additional lens on the interview data, helping us identify patterns that could be compared with our initial manual analysis. Its value was not in replacing interpretation, but in strengthening the transparency and robustness of our analytic decisions. At the same time, our study underlines that such tools must be used with careful human oversight — researchers remain responsible for interpretation, context, and the theoretical framing of the findings.”

For teacher educators, the article suggests that AI competence cannot be reduced to tool familiarity. Future teachers need chances to test GenAI, examine its limitations, discuss ethics and privacy, and connect AI use to theories of learning, feedback, assessment, and professional judgement.

For researchers, the study offers a useful model for combining survey evidence with interview data. The survey shows the scale of attitudes and beliefs, while the interviews reveal how experience changes the way participants talk about usefulness, risk, and human judgement.

Professor Gamlem offered a takeaway for readers considering the future of AI in classrooms: “The key takeaway is that generative AI should neither be uncritically embraced nor dismissed in education. What matters is preparing teachers who can engage with these tools in informed, reflective, and pedagogically grounded ways—understanding both their potential and their limitations.”

Taken together, the article invites a careful middle path: neither rejecting GenAI outright nor treating it as a substitute for educational expertise, but preparing teachers to use it with confidence, evidence, and care.