Generative AI dependency: the emerging academic crisis and its impact on student performance—a case study of a university in Zimbabwe

This study investigates the emerging phenomenon of generative artificial intelligence (AI) addiction among university students in Zimbabwe and its effects on academic performance. Through a mixed-methods approach combining surveys, interviews, usage tracking and academic performance data from 248 undergraduate students at a major Zimbabwean university, we found that 32.7% of participants demonstrated addictive patterns in their generative AI usage. Students exhibiting addiction showed distinctive behaviors including compulsive checking (averaging 18.3 daily AI interactions), failed attempts to reduce usage (reported by 65.8%) and continued reliance despite recognizing negative academic consequences. Statistical analysis revealed significant negative correlations between addiction severity and academic metrics, with heavily addicted students showing a mean GPA deficit of 0.41 points compared to non-addicted peers. Qualitative findings identified unique contextual factors in the Zimbabwean setting, including infrastructure limitations creating binge usage patterns, economic pressures increasing dependency and institutional resource constraints amplifying generative AI’s appeal. Multiple regression analysis identified three primary academic impact pathways: critical thinking atrophy, diminished writing skill development and reduced content knowledge acquisition. These findings contribute to understanding an emerging technological crisis in developing contexts and suggest targeted intervention approaches for Zimbabwe’s higher education environment.