Analyzing How AI Can Both Exacerbate and Help Overcome Digital Inequalities in Education, Especially in Underserved Regions

Abstract

This study examines the effects of AI-based learning aids on students' learning achievements, the association between the access to AI and digital disparities, and the efficacy of AI-based interventions in enhancing students' and teachers' digital literacy in deprived areas. Statistical comparisons based on data from 225 participants indicated no substantial effects of AI-based tools on learning achievements, as the association was very poor (R² = 0.002, p = 0.458). Likewise, the relationship between AI accessibility and digital inequalities was weak and not statistically significant (-0.019, p = 0.775), signifying that AI cannot on its own close learning gaps. Additionally, AI interventions were not effective in improving digital literacy, as evidenced by low mean values of the survey answers. These results are consistent with the earlier research highlighting the importance of supportive infrastructure, teacher education, and policy interventions to achieve the maximum potential benefits of AI in education. The research points towards the need to supplement AI-powered tools with holistic strategies that enhance digital access, provide necessary skills to teachers, and bring equitable implementation into education systems. Future studies must investigate intervention-based methods, long-term effects, and policy models to maximize the efficiency of AI in advancing digital literacy and diminishing educational disparities.