ABSTRACT Generative artificial intelligence (GAI) presents higher education (HE) with both significant risks and important opportunities for advancing decolonisation. This paper develops a systems thinking (ST) framework for the ethical and responsible integration of GAI in HE, bringing together ethical AI design, adaptive pedagogy and collaborative governance to support epistemic justice, mitigate algorithmic bias and enable inclusive curriculum transformation. Drawing on a qualitative document analysis of eight recent case studies, synthesised thematically and mapped visually, the study demonstrates how GAI can either reproduce colonial epistemic hierarchies or support the co‐creation of plural and culturally grounded knowledge systems. Using the ST Iceberg Model as an analytical structure, the paper connects observable AI practices to deeper recurring patterns, institutional structures and underlying mental models that shape educational outcomes. It further identifies concrete strategies for mental model transformation, including double‐loop learning, boundary critique and the use of targeted leverage points. Comparative insights from adjacent domains such as FinTech, Industry 4.0 and quantum systems are used to illustrate why socio‐technical interdependencies and governance choices are central to equitable and sustainable change. The paper concludes with practical recommendations for educators and policymakers focused on co‐design, ethical data governance, multilingual content development and capacity building. When guided by decolonial principles and systems‐aware implementation, GAI has the potential to support the reimagining of HE as a space of epistemic plurality, cultural relevance and collective empowerment.
Mohammed Albakri (Fri,) studied this question.