Fostering critical AI competency: a structural equation model of pre-service teachers’ trust and actual AI use in higher education

In recent years, education has experienced rapid change due to the rise of artificial intelligence (AI). This study analyzed how twelve interconnected factors, based on technology acceptance and trust theories, influence trust in AI for learning (TL) and actual AI use (AU) among pre-service teachers in Thailand. Using a quantitative design, with data collected from 260 pre-service teachers through purposive sampling based on prior AI experience. A 60-item, 5-point Likert-scale questionnaire, validated through pilot testing and internal consistency analysis (α=0.82–0.91). Data was analyzed using structural equation modeling (SEM) with maximum likelihood estimation. The model showed very good fit (χ²/df=1.601, root mean square error of approximation (RMSEA)=0.049) and explained 90.80% of behavioral intention (BI) and 71.20% of AU. Results indicated that cognitive load regulation (CLR) was the strongest predictor of TL (β=0.786, p