Why human leadership still matters under AI augmentation: PLS-SEM and machine-learning evidence from higher education in a global south context

Abstract

As artificial intelligence reshapes knowledge work, an important question emerges: Does human leadership still matter? This study examines how transformational leadership (TL) influences innovative work behavior in AI-augmented higher education through motivation to learn (MTL), while considering the moderating roles of perceived innovation climate (PIC) and AI-augmented job complexity (AAJC). We combined Partial Least Squares Structural Equation Modelling (PLS-SEM), Random Forest Regression, and SHapley Additive exPlanations (SHAP) to triangulate explanatory and predictive evidence. Survey data from 300 university employees in Vietnam were analyzed. During AI augmentation, transformational leadership still enhances innovative work behavior, directly and indirectly, by enhancing motivation to learn. Higher AI-augmented job complexity amplifies rather than replaces these human-driven pathways. Perceived innovation climate further reinforces the leadership–innovation relationship. The findings of this study challenge the assumption that advanced technology alone drives innovation. Instead, AI augmentation appears to amplify the importance of transformational leadership and continuous learning, particularly under conditions of higher AI-augmented job complexity. The study therefore redirects attention from technological capability toward human and organizational readiness in AI-driven transformation, especially within knowledge-intensive organizations in the Global South.