Abstract
Artificial intelligence is transforming how educational achievement is measured, interpreted, and applied. Tools such as automated scoring, adaptive testing, behavioural monitoring, and predictive models are now integral to assessment systems. While these technologies offer benefits like personalised learning, reduced administrative workload, and deeper performance insights, they also raise critical ethical concerns. This chapter explores these issues through three lenses: algorithmic fairness, structural equity, and institutional governance. It highlights how AI can both mitigate and reinforce inequalities depending on its design and implementation. Drawing on interdisciplinary research, the chapter argues that responsible AI integration requires more than technical accuracy—it demands a reorientation toward educational values, inclusive decision-making, and strong accountability frameworks. It concludes by proposing a five-pillar model to support ethical, equitable, and pedagogically sound use of AI in educational assessment.

This publication has 16 references indexed in Scilit: