Artificial Intelligence in Educational Assessment: Implications for Equity, Epistemology, and Development in Fiji and the Pacific

Artificial intelligence (AI) is reshaping the landscape of educational assessment, offering opportunities to move beyond standardized examinations toward adaptive, personalized, and competency-based approaches. Globally, AI-driven assessments provide real-time feedback, enable multimodal evaluation, and foster inclusivity for learners with diverse needs (Luckin et al., 2016; Holmes et al., 2022). For Fiji and the wider Pacific, these developments hold transformative potential in addressing persistent challenges such as inequitable access, limited teacher capacity, and reliance on narrow, exam-oriented systems. At the same time, the integration of AI into assessment raises critical questions about epistemology and the future of Pacific knowledge systems. Traditional learning in Fiji and the Pacific is grounded in communal, relational, and holistic epistemologies (Nabobo-Baba, 2006; Thaman, 2003), which may be marginalized if AI-driven assessment frameworks privilege Western, data-centric approaches. This paper critically examines the dual role of AI in assessment: as a tool for advancing equity, inclusivity, and educational development, and as a disruptive force that risks exacerbating the digital divide and eroding cultural epistemologies. It argues that for AI to contribute meaningfully to educational transformation in Fiji and the Pacific, policies and practices must prioritize cultural relevance, contextual adaptation, and equity-driven implementation. By aligning AI innovations with Pacific pedagogical values and development goals, educational assessment can evolve into a more inclusive, sustainable, and culturally grounded practice in the 21st century.