Iterative development of an AI intervention for pre-service physics teachers from a Vygotskian perspective

While artificial intelligence (AI) is transforming many sectors, its integration into pre-service teacher education in higher education remains limited. This study investigates the iterative development and effects of a concise, two-session educational intervention designed to foster AI literacy among pre-service physics teachers. Following a design-based research approach, the intervention was implemented in two iterations at the University of Cologne (n = 31 across two cohorts). Structured according to the 5E instructional model, the intervention required students to use generative AI tools as didactic instruments to create lesson plans and reflect on their usage. AI literacy was measured using a validated 30-item test, while attitudes toward AI were assessed via a 4-point Likert survey. Results indicate only small, non-significant increases in overall AI literacy, with selective gains observed in competencies explicitly supported by hands-on activities and targeted scaffolding. However, attitudinal measures demonstrated that even brief interventions can strengthen participants’ openness toward AI and their perceived preparedness to use AI tools in teaching. Additionally, the iterative comparison highlighted format-sensitive effects. These findings suggest that while short design-based interventions can selectively activate elements of AI literacy and foster professional confidence, they are insufficient for broader skill Acquisition. Consequently, more sustained, context-rich engagements are likely required to achieve comprehensive and durable AI literacy development in pre-service teacher education.