Training in the Age of Generative AI

This chapter examines the determinants of generative artificial intelligence use by higher education teachers in Morocco utilizing a mixed methodological approach based on multiple linear regression, machine learning models (Random Forest, XGBoost, SVM), and partial least squares structural equations (PLS-SEM). Based on a survey of 252 teachers using a three-part questionnaire, the findings show three primary and converging levers to the adoption of these technologies: participation in continuing education, prior digital experience, and perceived institutional support. The AI models show high predictive performance (F1 > 0.80; AUC > 0.85), and the PLS-SEM model substantiates the theoretical consistency of the conceptual framework (R2 = 0.61). Socio-demographic variables contributed little. This methodological triangulation allows for the combination of statistical robustness and explanatory richness, and provides concrete recommendations to support pedagogical transformation in the age of generative AI.