Adaptive learning through Artificial Intelligence in Higher Education

Abstract

The research analyzed the current state of adaptive learning through artificial intelligence in higher education, aiming to examine theoretical frameworks, practical implementations, empirical benefits, and emerging challenges of these technological systems. A systematic review methodology with narrative synthesis was employed, analyzing nine specific academic references from the 2020–2025 period, complemented by a systematic search in academic databases. The results showed that most studies reported positive effects of adaptive learning, with statistically significant improvements in academic performance, student perception, and the accuracy of automated responses in systems. Convergent theoretical frameworks were identified, trending toward hybrid constructivist–AI models, specifically the GenAI-TPACK framework as the dominant approach for responsible integration. Technical architecture followed the four classic components, enhanced with natural language processing and advanced predictive analytics, while successful implementation strategies combined intensive teacher training with collaborative curriculum redesign and phased, gradual deployment. The main challenges include ethical concerns regarding privacy—where educators expressed significant apprehension over cyberattacks—high implementation costs that can exceed two million dollars, technical complexity identified by organizations as the primary barrier, institutional resistance linked to insufficient teacher training, and the persistent digital divide that may exacerbate existing inequalities between wealthy and under-resourced institutions. The findings confirm that AI-powered adaptive learning represents a fundamental transformation capable of democratizing access to high-quality personalized education in higher education institutions.