Generative Artificial Intelligence integration in higher education: A constructivist learning theory approach

• Most faculties use GAI for information search, followed by research activities. • Faculties’ demographic variables are not alone sufficient to explain their GAI use. • GAI is a supportive tool, not a replacement. • Need to revise the curriculum, not resist the technology. • Women are often left behind in tech adaptation. • Faculties depend on their personal strategies in the absence of standardized institutional guidelines. The integration of Generative Artificial Intelligence (GAI) in higher education presents transformative opportunities for teaching, learning, and research, particularly in resource-constrained developing countries like Bangladesh. This study investigates the potential of GAI adoption in Bangladeshi universities through the lens of Constructivist Learning Theory (CLT). Employing a mixed-method approach combining surveys (n=255 faculty members) and Focus Group Discussions (FGDs), the research explores factors influencing GAI adoption, including competency, domain relevance, and institutional support. The research employed Structural Equation Modeling (SEM) using SmartPLS to examine the relationships among key constructs. The findings highlight that attitude towards GAI use is a strong predictor of the extent of GAI use (β = 0.556, p < 0.001), influenced significantly by users’ competency, domain-specific technology relevance, and students’ technology use provision. Among the predictors of attitude towards GAI use, domain-specific technology relevance exhibited the strongest effect (β = 0.413, t = 6.561, p < 0.001). However, subjective norms and institutional technology use provision are not found to be significant in the context. While students' technology use provision showed a positive and significant contribution, the findings of thematic analysis indicated the teachers’ strong concern regarding technology use. Post hoc analysis revealed that the control variables age, gender, education, and academic role had negligible effects. The study recommends enhancing university capacities through training, integrating GAI into pedagogy, and promoting ethical guidelines and support for both students and teachers. Future research should examine the longitudinal effects of GAI on teaching outcomes and institutional readiness in various contexts.