GENERATIVE AI USE AND SKILL DEVELOPMENT: MULTIGROUP ANALYSIS ACROSS PAKISTANI HIGHER EDUCATION STAKEHOLDERS

Abstract

The present study focuses on the impact of generative artificial intelligence (GenAI) on the process of skill development within higher education institutions by employing a serial mediation framework involving AI use, information literacy, cognitive engagement, and skill development across multiple stakeholders in Pakistan. The study is based on the Technology Acceptance Model (TAM) and Constructivist Learning Theory (CLT), highlighting the critical importance of cognitive activity over GenAI exposure in skill development. A quantitative, cross-sectional research design with stratified random sampling of 669 participants was conducted, including students, faculty, and administration working at higher education institutions in Pakistan. PLS-SEM and MGA were employed to analyze the data, with a total of 5,000 bootstrap samples being considered. Results indicated that AI use had a significant impact on information literacy (β = 0.381, p < 0.001), which facilitated cognitive readiness (β = 0.479, p < 0.001) for skill building (β = 0.369, p < 0.001). It was found that about 94% of the total influence of AI use on skill building was mediated through information literacy and cognitive readiness. The multi-group analysis identified stakeholder differences such that administrative staff had more literacy effects, faculty members had more skill-building impacts directly, while students had less literacy but greater skill-building effects via cognition. Findings show that information literacy and cognitive readiness were found to be significant compared to AI use. The theoretical contribution of this study lies in the incorporation of TAM and constructivist learning into serial mediations while practical contributions lie in emphasizing the need for targeted digital literacy programs for effective adoption of AI.