This study aims to explore decolonised strategies for challenging hegemonic assessment practices in teacher education, which often equate academic quality with dominant English-language writing conventions. It confronts the perception that effective assessment inherently privileges specific linguistic norms, arguing instead for a social justice approach that reconceptualises evaluation to empower all students. The purpose is to propose how Artificial Intelligence (AI) can be harnessed within a decolonised framework to develop inclusive assessment methods that assess for learning at the Department of Educational Foundations ‘ B.Ed. Honours programme at the University of South Africa. The methodology involved a systematic literature review across major academic databases (ERIC, Scopus, Web of Science, Google Scholar) and specialised journals. Initial searches using terms related to decolonised pedagogy, inclusive assessment, and AI in education yielded approximately 100 papers. After screening titles, abstracts, and full texts for theoretical depth and conceptual relevance, 24 publications were selected for in-depth analysis based on strict criteria of relevance, rigour, and contribution to the synthesis of a decolonised AI approach. The discussion synthesises the literature to argue that AI strategies, when guided by a decolonial ethos, may provide innovative models for supporting academic writing and reframing assessment. This approach may help dismantle repressive structures by moving away from a deficit model and towards one that values diverse student voices and backgrounds. The study recommends that academics intentionally integrate prescribed AI tools into Honours-level assessments to promote learning and equity. The conclusion asserts that a decolonised approach to assessment, augmented by AI, may transform standard practices to advantage historically underrepresented students, ultimately aligning assessment with the goals of social justice and inclusive education.