Artificial Intelligence for improving e-learning: Opportunities, limits, and design principles

Abstract

Artificial intelligence is changing how students learn online. Evidence from classrooms remains limited and sometimes inconsistent. This study tests AI support within real e-learning courses. We used a quasi-experimental design across two undergraduate courses. One section received AI tools embedded in the LMS. Another section used standard materials without AI features. The Courses were Management Information Systems at a public university. Students accessed materials using laptops and mobile phones. The chatbot explained concepts and answered questions. Adaptive quizzes adjusted difficulty using aligned item banks. Recommendations surfaced videos and readings from the course. Control sections received identical content without automation. We balanced instructors, topics, and assessments across sections. The intervention lasted eight teaching weeks. Intervention fidelity was monitored through weekly instructor checklists. We measured learning gains with aligned pretests and posttests. Engagement came from system logs of sessions and interactions. A validated survey captured satisfaction and perceived usefulness. All students provided consent under approved ethics procedures. We examined outcomes using mixed-effects models controlling for relevant variables. AI sections outperformed controls in posttests after controlling for baseline. More sessions and faster feedback cycles resulted in higher engagement. Students reported higher satisfaction and greater sense of supervision. Instructors reported reduced grading load for routine tasks. We observed no clear rise in integrity violations. Adoption required onboarding time and careful data governance. The Short duration and single institution limit generalizability. AI can augment teaching when aligned with course pedagogy. Future work should test longer deployments across diverse contexts. We provide practical guidelines for responsible classroom adoption.

Keywords: Artificial intelligence, E-learning, Smart education, AI applications, Modern technology teaching, Adaptive Learning.