TRANSFORMING HIGHER EDUCATION WITH AI: A FRAMEWORK FOR STUDENT LEARNING AND INSTITUTIONAL READINESS

The research synthesizes and tests a framework that connects individual student traits and institutional characteristics to academic performance outcomes in the context of AI-augmented higher education. We hypothesize, based on previous investigations, that the independent variables AI Self-Efficacy (AISE), Digital Literacy (DL), and Institutional Readiness (IR) have a bearing on students’ engagement and on the adoption of academic tools, which then determines Academic Performance (AP), Employability Readiness (ER), and Creativity / Innovation (CR). We collected data from 600 university students, applied validated scales to measure the constructs, and analyzed the results through Partial Least Squares Structural Equation Modelling (PLS-SEM). The measurement model demonstrated strong reliability (Cronbach’s α > .80 for all constructs) and convergent validity (AVE > .50). Examining the structural model, the relationships from AISE, DL, and IR to AI Tool Adoption (AIA) and Learning Engagement (LE) were all positive and significant (β ≈ .30-.50, p < .001). The two mediators (AIA and LE) were positive predictors of AP, ER, and CR (β ≈ .25-.45, p < .01), and they explained substantial variance (e.g., R²_AP ≈ .50, R²_ER ≈ .43, R²_CR ≈ .38). Based on this, we might cultivate students' confidence in AI and their digital skills, as well as provide them with supportive surroundings, to improve educational outcomes and spur innovation. We might do this by investing in infrastructure, as well as in training programs, and AI pedagogical and policy implications are discussed. Limitations (such as the cross-sectional design) and suggestions for future studies are also mentioned.