Drawing the line: A Qualitative Study on Students' Anti-Consumption of Large Language Models

Sammanfattning

This bachelor thesis explores the underlying values that shape students' boundaries and their decisions to refrain from using Large Language Models (LLMs) in their academic work. While previous research has primarily focused on the drivers of adoption and usage of generative AI, this study highlights the understudied phenomenon of meaningful non-use and the subjective processes that lead to limited usage.Using a qualitative method based on semi-structured interviews with business and economy students, the phenomenon is analyzed through the lens of anti-consumption. The study utilizes Lee et al.’s (2009) framework on brand avoidance, with a specific focus on the dimensions of identity and morality, to understand how students navigate the tension between the benefits of efficiency and the negative symbolic associations of AI use.The findings indicate that students' navigation of LLM tools is characterized by an internal identity conflict and a moral balancing act. The decision to refrain from LLMs is often contextual and linked to the saliency of the student's identities; in contexts where human creation and genuine learning are highly valued, avoidance is activated to protect the "desired self" and to distance the individual from an "undesired self" associated with laziness or incompetence. The study contributes to a deeper understanding of how students create meaning around anti-consumption of LLMs and how personal values function as internal mechanisms of control when external academic regulations are perceived as ambiguous.