Ethical challenges and regulatory aspects of artificial intelligence use in academic performance assessment systems

Abstract

The integration of artificial intelligence into educational processes, particularly in systems for assessing learners’ academic performance, is accompanied not only by increased efficiency in processing educational data but also by heightened ethical and regulatory controversies. The automation of assessment-related decision-making transforms traditional approaches to academic responsibility, transparency, and fairness, thereby necessitating a comprehensive scholarly examination of these processes. The purpose of the article is to investigate the ethical challenges and regulatory considerations of using artificial intelligence in systems for assessing learners’ academic performance, and to develop conceptual approaches to its responsible implementation within the educational environment. Methods. The study employs methods of systemic and structural-logical analysis, comparative analysis of regulatory and legal approaches, generalization of scholarly positions on the ethics of algorithmic decision-making, and conceptual modeling to develop an integrated model for the use of artificial intelligence in learning outcomes assessment. Results. Key ethical challenges arising from the application of artificial intelligence in academic performance assessment are systematized, including algorithmic bias, lack of transparency in assessment decisions, data confidentiality breaches, and the erosion of pedagogical responsibility. The main regulatory approaches to the use of artificial intelligence in education at international and institutional levels are analyzed. A conceptual model of responsible artificial intelligence use is proposed, combining algorithmic support with mandatory human oversight. Conclusions. The expediency of implementing a human-centered approach to the use of artificial intelligence in academic performance assessment systems is substantiated. It is demonstrated that the effective application of algorithmic tools is possible only under conditions of adherence to ethical principles, regulatory compliance, and the preservation of the instructor’s key role in final decision-making.