The Effectiveness of Generative AI as a Learning Support Tool: Evidence from Higher Education in Indonesia

Abstract

Generative Artificial Intelligence (GenAI) has become an increasingly relevant learning support technology in higher education. This study examines its effectiveness in supporting learning among higher education students in Indonesia by synthesizing recent empirical and methodological evidence. A descriptive literature review was conducted using five relevant journal articles published between 2025 and 2026. The reviewed studies addressed perceived AI effectiveness in online learning, personalized learning, student engagement, digital and AI literacy, and machine learning-based analysis of learning effectiveness. The synthesis indicates that GenAI can support material comprehension, flexible individual learning, and personalized learning recommendations. One reviewed study reported an overall perceived effectiveness score of 3.44, with material comprehension reaching 3.55. Another reported a 15.03% increase in average academic performance among students using an AI-based learning analytics tool, compared with 4.63% among non-users. However, the evidence also shows that frequent AI use alone does not guarantee stronger engagement or learning outcomes. Digital and AI literacy, active student participation, lecturer guidance, institutional support, and responsible use influence the value of GenAI. Key risks include inaccurate information, cognitive dependency, reduced critical thinking, privacy and security concerns, infrastructure limitations, and academic integrity. Therefore, GenAI is best positioned as an augmentative learning support tool rather than a replacement for lecturers or independent learning. Future research should use longitudinal or experimental designs and objective learning indicators to establish stronger evidence of its effects.