<![CDATA[The use of Artificial Intelligence (AI) in educational evaluation optimizes learning outcomes. This research seeks to address the advantages and difficulties of implementing AI within academic evaluation frameworks with particular emphasis on the algorithmic bias problem and its implications for fairness in education. The absence of a thorough grasp of algorithmic bias, particularly how it can be utilized as a weapon against equitable education, reveals an important gap. We conduct a Systematic Literature Review (SLR) and bibliometric analysis on 121 articles sourced from Scopus published between 2021 to 2025 to trace the trends and examine the impacts and biases of AI on grading systems. The data demonstrates a significant increase in publications beginning 2018, concentrating on topics such as educational applications of AI, automated grading systems, and machine learning. The findings further indicate that though AI improves efficiency and consistency of the evaluations, it heightens the chances of biased outcomes because of non-diverse training data, prejudiced developers, and socio-cultural frameworks that could worsen the situation for already marginalized learners. In summary, this study highlights the critical gaps in bias mitigation strategies arising from the lack of ethical design frameworks, antecedent-free algorithms, and educator prep courses aimed at combating bias. These outcomes serve as benchmarks for the creation of more reliable and comprehensive AI systems for assessments and shift subsequent investigations to focus validation on different cultures and the incorporation of just AI design paradigms]]>
Copyrights © 2025
