The Concern Over Brain Rot from Generative AI Use Among Preservice Teachers: A UTAUT Approach

Abstract

Purpose – The increasing use of generative AI on campus has raised concerns about a potential decline in students’ critical thinking skills. While the UTAUT theory is widely used to examine technology adoption, its relationship with the phenomenon of brain rot remains underexplored, particularly among preservice teachers. This study aims to analyze the factors associated with preservice teachers’ intention to use generative AI within the UTAUT framework, as well as to examine its association with tendencies toward brain rot.Method – A quantitative cross-sectional design was conducted with 243 preservice teachers from Universitas Negeri Makassar. Data were collected via a validated 30 item questionnaire and analyzed using Partial Least Squares Structural Equation Modeling (PLS-SEM) to examine the relationships between technology adoption constructs and brain rot tendencies.Findings – Social influence was the only significant predictor of behavioral intention to use AI (β = 0.269, p = 0.002). Behavioral intention, in turn, showed a strong positive association with brain rot tendencies (β = 0.817, p < 0.001), explaining 66.7% of the variance (R² = 0.667). Other UTAUT constructs, including performance expectancy and effort expectancy, were not significant predictors. However, given the cross-sectional design, these findings reflect statistical associations rather than causal relationships.Research Implication : Socially driven AI adoption is strongly linked to cognitive passivity, highlighting the need to extend UTAUT with cognitive risk factors and rethink how technology use impacts higher-order thinking.Conclusion – This study indicates that the adoption of AI among preservice teachers is associated with perceptions of declining cognitive abilities. These findings highlight the importance of promoting critical AI literacy and developing assessment approaches that emphasize deep cognitive engagement. Future research is recommended to employ longitudinal designs or incorporate control variables such as digital self-efficacy.