Navigating the AI Divide: A Comparative Analysis of Institutional Readiness and Generative AI Policies in Australian and Pakistani Higher Education

Abstract

The rapid proliferation of Generative Artificial Intelligence (GenAI) has fundamentally disrupted global higher education, prompting a rush to establish institutional governance. However, prevailing discourse often frames this integration as a uniform technological transition, obscuring the profound socio-economic divides that dictate institutional responses in the Global South. This study conducts a qualitative comparative document analysis of primary AI policy frameworks to investigate how macro-economic resource disparities influence policy maturity. Utilizing stratified purposive sampling, this research contrasts national regulatory bodies and institutional tiers-ranging from elite to public universities—within a high-resource ecosystem (Australia) and a resource-constrained environment (Pakistan). Findings reveal that high-resource environments, guided by bodies like TEQSA, operate with universal infrastructural readiness, allowing them to confidently micro-manage AI as an everyday pedagogical tool. Furthermore, elite Pakistani institutions possess the autonomous capital to bypass domestic lags, achieving regulatory convergence with Australia’s elite Group of Eight (Go8). However, major public institutions in resource-constrained environments remain trapped in a "technical paradox." Forced to centralize scarce funding into specialized tech-hubs, these institutions successfully train AI engineers but suffer a severe pedagogical policy vacuum, resulting in siloed human capital and high academic friction for non-technical students. The study concludes that navigating the AI divide requires a departure from Western-centric regulatory assumptions. It recommends that developing educational markets transition from aspirational mandates to highly prescriptive frameworks, emphasizing cross-faculty democratization, low-resource assessment redesign, and the deployment of localized open-source models to prevent the acceleration of a domestic and international digital divide.