Capability-based training framework for generative AI in higher education

Abstract

Integrating generative artificial intelligence (GenAI) in higher education (HE) requires educators to develop new competencies. However, while GenAI holds transformative potential for education, research on the competencies needed for its responsible and effective use remains limited. This study employs a mixed framework analysis method, combining quantitative and qualitative analysis to identify key competencies essential for HE teachers. The research began with a bibliometric analysis of 1,737 documents from Scopus and proceeded with an in-depth analysis of 14 peer-reviewed articles. Using a chain-of-thought (CoT) prompting approach, the analysis integrates a human-GenAI collaboration to identify patterns in existing competency frameworks and empirical publications, aiming to classify and define competencies. The findings reveal that while AI literacy and ethical awareness are frequently mentioned, there is no unified competency framework addressing the pedagogical and technical dimensions of GenAI integration. The FAM process resulted in the identification of three key domains of competencies and a set of 16 competencies. The results highlight the need for a structured, yet flexible competency model tailored to educators. Future research should focus on empirical validation and the development of professional development programs to bridge the identified gaps.