Presentation of A GPT is a GPT, but can it Equalise? A Quantitative Study of Adoption, Inclusion, and Engagement from Students' Perceptions and Reported Use of Generative AI Tools

This study investigates how student engagement with Generative Artificial Intelligence (GAI) tools is influenced by perceived usefulness, socioeconomic status (SES), and institutional support within an African research-intensive university. As GAI becomes increasingly integrated into higher education, the question of equitable adoption and meaningful engagement remains underexamined, particularly in postdigital contexts characterised by persistent digital and social asymmetries. To address this gap, the study draws on an integrated theoretical framework that synthesises the Technology Acceptance Model, Digital Divide Theory, and Self-Efficacy Theory, articulated here as the Adoption, Inclusion, and Engagement (AIE) framework. Employing a deductive, positivist design, the study administered a structured survey to 581 students across diverse faculties. The analysis, based on non-parametric statistical tests, revealed that perceived usefulness significantly correlates with GAI engagement (Spearman’s ρ = 0.524, p < 0.001), SES significantly affects usage frequency (U = 26,530.5, p = 0.0278), and perceived institutional support moderates levels of engagement (H = 7.30, p = 0.0259). These findings demonstrate that student engagement is not merely a function of technological access, but is deeply shaped by students’ perceptions of value, confidence in their critical digital competence, and their sense of institutional inclusion. This study contributes to the evolving field of GAI adoption in higher education by proposing the AIE model and advocating for equity-focused digital strategies that support meaningful participation for all students, particularly within structurally unequal but pedagogically rich African university contexts. Research affilliated to the GAI Project at Wits, this slideshow was created and developed by Malcolm Weaich under the supervision of Prof. Kershree Padayachee at The University of the Witwatersrand. The full article linked to the submission is to be submitted to the Journal of Postdigital Science and Education and this slideshow was presented at the Thirty-Second International Conference on Learning on Human Learning and Machine Learning, Challenges and Opportunities for Artificial Intelligence in Education by Malcolm Weaich on the 10th of July 2025. The conference date was the 8-10 July 2025 at the University of Granada, Granada, Spain. Malcolm Weaich was subsequently awarded the Emerging Scholar Award at the conference for the presentation of the research. Copywrite Licence: CC BY-SA 4.0 DEED.