Abstract:

Universities are integrating artificial intelligence (AI) into student-facing decision processes faster than they are working through what that means for the students affected by it. Admissions screening, scholarship allocation, academic advising, assignment marking, and early-warning systems are increasingly shaped by algorithmic tools that promise analytical rigor, objectiveness, efficiency, and the potential to reduce administrative headcount. Discussion of these systems has tended to center on their accuracy, sophistication, and technical readiness for high-stakes use. Much less attention has been paid to a different question that may matter just as much in practice: Do students experience AI-mediated decision-making as fair, understandable, and worthy of their trust?

That question is especially important in internationally diverse educational institutions because students from different parts of the world approach institutional decision-making with differing understandings of how those with power over them should operate, whether they are owed explanations for decisions that affect them, or what it means for an outcome to be fair. Simply put, it is not safe to assume that students will interpret the same decision-making process and outcome in the same way. Whereas one student might judge it appropriately rigorous, it may be disturbingly opaque to another.

This paper draws from organizational justice and cross-cultural management scholarship to examine how universities can use AI-supported academic decision systems without undermining student trust or, under the right conditions, potentially strengthen it. Organizational justice research has long demonstrated that people weigh consequential decisions not only by their outcomes but also by the fairness of the processes, explanations, and interpersonal treatment they receive as those outcomes are decided. Cross-cultural management research shows that such judgments are not made from a universal standpoint but are shaped by culturally-encoded assumptions about authority, legitimacy, and what treatment and explanations institutions owe their members. Together, these literatures offer a useful framework for understanding how students are likely to interpret AI-supported academic decisions in internationally diverse institutions.

Ensuring fair and accurate criteria for AI-mediated decisions is vital, but that is not enough on its own. Institutions must also be attentive to how they present the systems, explain their role, and build governance around their use. When institutions fail at this, trust will erode. Such failures may take the form of inaccessible decision logic, advisors deferring to the system rather than interpreting it, not offering students a meaningful route to raise concerns or seek review, or ignoring students' differing perceptions of what fairness is and what explanations are required.

To help avoid these pitfalls, this paper proposes design principles centered on transparency, visible human oversight, accessible appeals, and culturally responsive communication that should be built into AI-supported academic decision systems from the outset. The long-term legitimacy of AI in higher education will not be determined by technical performance alone. It also depends on students experiencing these systems as fair, accountable, and consistent with what they reasonably expect an academic institution to be.

Keywords:

Higher education, algorithmic decision-making, student trust, organizational justice, AI governance, institutional trust, cross-cultural management.