Abstract

The integration of artificial intelligence (AI)-assisted tools in classroom pedagogy has gained significant momentum globally, yet critical questions surrounding reliability, ethical integrity, and instructional validity remain underexplored in SubSaharan African contexts. This study presents a critical appraisal of AI-assisted pedagogical tools employed in secondary and tertiary educational institutions in Ondo State, Nigeria. Employing a mixed-methods research design, the study collected data from 320 teachers and 480 students across 24 purposively selected institutions in Ondo State. Structured questionnaires, in-depth interviews, and classroom observation protocols were utilised as instruments. The study adopted Technology Acceptance Model (TAM) and Constructivist Learning Theory as theoretical anchors. Quantitative data were analysed using descriptive statistics, Pearson correlation, and multiple regression analysis, while qualitative data underwent thematic analysis. Findings reveal that while AI-assisted tools significantly enhance instructional delivery (r = 0.712, p < .001), critical deficiencies in reliability (mean = 3.21/5.00), ethical safeguards (mean = 2.87/5.00), and curricular alignment (mean = 3.04/5.00) constrain their pedagogical effectiveness. The study identifies teacher digital competence, infrastructure adequacy, and institutional policy frameworks as significant predictors of AI tool effectiveness. Implications for curriculum developers, institutional administrators, and education policymakers in Ondo State and comparable contexts are discussed. The study recommends the establishment of state-level AI governance frameworks and the development of context-sensitive AI integration protocols