Generative AI and digital neocolonialism in global education: Towards an equitable framework

Abstract

As generative artificial intelligence (GenAI) becomes increasingly embedded in education systems worldwide, urgent questions arise concerning whose knowledge these technologies elevate and whose they marginalize. This study adopts a twofold critical–constructive approach to examine GenAI’s role in reproducing epistemic hierarchies and to advance pathways toward more equitable use in education. Using a critical constructive qualitative design, we first conducted zero-shot prompt testing with ChatGPT-4 Turbo and Gemini 1.5 models across contexts in the Global North and Global South. The models responses were documented in real time and analyzed through a critical interpretive lens to surface patterns associated with digital neocolonialism. The critical phase of the study identifies six interconnected dimensions through which GenAI sustains Western dominance in educational contexts: Western curriculum ideologies, cultural imperialism, pedagogical control, language marginalization, racial and ethnic underrepresentation, and access inequity. For instance, when Gemini was asked to identify the seasons in the United States and Ghana, it returned the same four-season framework for both contexts, reflecting Western climatological assumptions. Across other prompts, GenAI outputs relied on stereotypical imagery, assumed Western-centered instructional resources, limited Indigenous and local language support, and disproportionately represented Western racial identities. In addition, subscription-based pricing models create structural barriers, as educators and institutions in much of the Global South face disproportionate costs due to currency differences. Building directly on these findings, the constructive phase advances two mitigation pathways for equitable GenAI in education. The first pathway targets AI design, emphasizing liberatory design methods, foresight by design, and the decentralization of GenAI development to strengthen local participation and data sovereignty. The second operates at the pedagogical level, advancing a human-centric prompt engineering model that empowers educators to contextualize prompts, critically interrogate outputs, and exercise pedagogical agency. These pathways position GenAI not merely as a technological tool, but as a site of ethical, and culturally responsive education.

Ähnliche Arbeiten