From AI Adoption to Meaningful Learning: A Systematic Review of AI-Based Pedagogy, Student Engagement and Critical Thinking in Higher Education

Abstract

Artificial Intelligence (AI) has rapidly evolved from a specialized computational technology into an increasingly important component of higher education, influencing teaching, learning, assessment and academic support. Intelligent tutoring systems, adaptive learning platforms, learning analytics, automated feedback, conversational agents and generative AI are creating new possibilities for personalized and interactive learning. At the same time, the rapid integration of AI has raised important concerns regarding student engagement, critical thinking, academic independence, AI literacy, teacher competence, digital infrastructure and responsible use. Existing research provides evidence of considerable educational potential, but findings remain fragmented across AI technologies, pedagogical approaches, disciplines and learning outcomes. This review critically examines the emerging literature on AI-based pedagogy in higher education, with particular emphasis on its relationship with student engagement and critical thinking. It further investigates the roles of AI literacy, perceived usefulness, teacher competence, AI-related technological-pedagogical knowledge, digital infrastructure and institutional readiness in determining the effectiveness of AI-supported learning. The review is conceptually grounded in Artificial Intelligence in Education (AIED), the Technological Pedagogical Content Knowledge (TPACK) framework, AI literacy and technology-adoption perspectives. The conceptual focus is consistent with the original manuscript, which identified AI literacy, perceived usefulness and digital infrastructure as important antecedents of AI-based pedagogy and positioned student engagement and critical thinking as major educational outcomes. A structured literature-review methodology is adopted to identify, screen, organize and thematically synthesize relevant research, with particular attention to AI-enabled personalization, adaptive learning, intelligent tutoring, generative AI, student engagement, critical thinking, AI literacy and institutional readiness. The review distinguishes between technological interaction and meaningful cognitive engagement, recognizing that frequent AI use does not necessarily indicate deeper learning. Similarly, AI may support critical thinking when used for questioning, comparison, evaluation and reflection, but may weaken independent reasoning when students simply delegate cognitive tasks to AI. The review proposes an AI–Pedagogy–Engagement–Critical Thinking (AI-PECT) framework in which AI capabilities influence educational outcomes through pedagogical mediation and learner agency, while AI literacy, teacher competence, infrastructure, assessment design and ethical governance operate as important contextual conditions. The review argues that the educational value of AI depends less on the presence of AI technology itself and more on how deliberately, critically and responsibly it is integrated into teaching and learning. The findings provide a foundation for educators, institutions and policymakers seeking to move from simple AI adoption toward meaningful, human-centred and intellectually responsible learning.