"The teachers don't say anything, the students don't ask" : An exploratory study on how students and teachers balance and negotiate tensions in school-related AI practices in Swedish upper secondary schools

Abstract

Purpose: Teachers and students are seen to be insecure and have conflicting views around the rules of AI use in schools. However, when students and teachers are seen calling for clearer rules, it should be interpreted as a symptom of these insecurities and not a solution per se. To establish new practices, schools need to identify and understand the tensions that students and teachers experience in their current school-related AI practices. Because of the contextual nature of both AI use and rules negotiations, the sparse research literature on how rules for AI use is negotiated in secondary schools comes too short. By exploring the characteristics of AI use and how this use is understood and negotiated by students and teachers, this thesis aims to gain deeper insights into what counts as accepted and non-accepted use of AI for school-related work. These insights can be used by schools to establish trustworthy, transparent, and ethically sound policies for AI use. Theory: The study utilises Cultural-historical activity theory (CHAT) and the Theory of expansive learning (Engeström, 2014). This theory provides a framework for understanding individual and organisational change through the interplay between individuals and the collective, tools, and rules. Method: Data were collected from focus groups where Swedish upper secondary teachers and students were discussing examples of problematic AI use. Through qualitative content analysis of these discussions, tensions around what is seen as acceptable use of AI were identified and explored. Results: Three tensions were found to be present in the student–teacher discussions. First, students and teachers are calling for clear rules, but also to be allowed to make their own adoption of the rules. Second, evaluating expected outcome of AI use is important for assessing the acceptance, but users doubt their own AI competency. Third, both teacher and students acknowledge that AI use is interfering with the learning process, but they also search for ways to use AI to improve students' learning or teachers' workflow. Negotiations within these tensions are taking place on an acceptance scale ranging from the non-acceptable use of AI as a substitute to the generally accepted use of AI as an assistant. To make decisions on acceptable AI use, actors are seen to negotiate acceptance through weighing the intention behind and the expected outcome of AI use against the perceived rules.