Student Acceptance of AI-Driven Deep Learning Tools: The Influence of Institutional Support and Teacher Competency in Vocational Schools

This study examines students’ acceptance of AI-driven deep learning tools in vocational schools by extending the Technology Acceptance Model (TAM) with contextual factors, namely Institutional Support (IS) and Teacher Competency (TC). The purpose of this research is to clarify how perception-based variables (Perceived Usefulness—PU, Perceived Ease of Use-PEU) interact with pedagogical and institutional conditions in shaping students’ Behavioral Intention (BI) to use deep learning tools. A total of 241 vocational high school students participated in the study. Data were analyzed using Structural Equation Modeling–Partial Least Squares (SEM-PLS) to test both direct and indirect effects. The results indicate several unexpected patterns. Although PEU significantly increases both ATU (β = 0.496, p < .001) and BI (β = 0.405, p < .001), PU shows mixed effects: while it positively affects ATU (β = 0.351, p < .001), its direct effect on BI is negative (β = –0.247, p < .001). ATU also negatively predicts BI (β = –0.235, p < .001), suggesting that favorable attitudes alone do not guarantee students’ intention to adopt AI tools. In contrast, Teacher Competency emerges as the strongest predictor (β = 0.891, p < .001), followed by Institutional Support (β = 0.087, p = .008). Indirect analyses show that PEU and PU exert significant negative mediation through ATU toward BI (PEU → ATU → BI: β = –0.117; PU → ATU → BI: β = –0.083). These findings imply that the successful integration of AI-driven deep learning tools in vocational education does not rely solely on students’perceived usefulness or attitudes. Instead, competent teachers and supportive institutional environments play a far more decisive role in shaping adoption intention. The study highlights the need for targeted teacher training, stronger institutional readiness, and pedagogical strategies that move beyond perception-based acceptance to ensure meaningful and sustainable AI integration in vocational learning contexts.