Abstract:
Artificial intelligence (AI) is increasingly transforming higher education through
applications in teaching, assessment, research, and institutional management. However,
existing studies remain fragmented and often overlook governance, security risks, and
ethical implications. This study presents a systematic review of AI tools in higher education
from a security science perspective. Using PRISMA 2020 guidelines, peer-reviewed studies
published between 2020 and 2025 were analyzed through thematic synthesis. The findings
identify four major categories of AI tools: generative AI, learning analytics systems,
intelligent tutoring systems, and administrative decision-support tools. While these
technologies enhance efficiency and personalization, they introduce risks related to
academic integrity, data privacy, algorithmic opacity, and system dependency. To address
this gap, the study proposes a weighted Confidentiality–Integrity–Availability (CIA) model
for quantifying AI-related risks, alongside a governance framework for institutional risk
management. The results emphasize that effective AI adoption requires robust governance,
ethical safeguards, and human-centered oversight. The study contributes a structured and
measurable approach to evaluating AI risks in higher education systems.
