PROTECTING DIGITAL LEARNING WITH AI-DRIVEN CYBERSECURITY: ENSURING ACCESS TO QUALITY EDUCATION

Abstract

The rise of digital learning has significantly expanded educational access, but it has also introduced vulnerabilities to cyber
threats that disrupt educational continuity and compromise sensitive data. This research examines AI-driven cybersecurity
in safeguarding digital learning platforms with BETH 2021 datasets to apply the 2 models Isolation-Forest and SupportVector-Machine (SVM) models for anomaly detection and predictive security. Isolation-Forest achieved the maximum
overall act with an F1 score of 0.894 on the test set, demonstrating its capacity to balance precision and ideal for educationfocused cybersecurity where minimizing false alarms is essential. In contrast, SVM achieved perfect precision (1.0) but
showed limitations in recall and highlighted its use for high accuracy in low-risk scenarios. The case study on AI-enhanced
security in Learning-Management-Systems (LMS) further illustrates practical applications and demonstrates AI's role in
real-time threat detection in secure exam proctoring and data protection. These findings underscore the broader significance
of cybersecurity in promoting educational equity and protecting digital environments with uninterrupted access to learning.
Future research is recommended to explore adaptive AI, privacy-first innovations, and blockchain integration for more
resilient and inclusive digital learning ecosystems.