AI-Enabled Digital Victimisation and Digital-Safety Preparedness among Secondary School Students in Gusau Metropolis, Zamfara State, Nigeria

Abstract

Background: Generative AI has enabled digital victimisation via AI-generated messages, impersonation, manipulated images and voices, and automated harassment. The study investigated the digital victimisation and preparedness in digital safety of secondary school students in Gusau metropolis, Nigeria using Artificial Intelligence (AI) as the aid tool.

Methods: 846 schools analytical cross-sectional study focused on 36 secondary schools. A culturally adapted English-Hausa questionnaire data was collected that measures the four factors: digital victimisation, AI literacy, trusted-adult support, reporting readiness, and scenario-based preparedness. School clustering was accounted for using two-level mixed-effects models.

Results: 37.8% stated that they had experienced AI-enable digital victimisation, 14.7% had experienced cyberbullying perpetrated by others and 8.5% had suffered a single incident of severe digital victimisation. Covariate adjusted, frequency of victimisation was negatively associated with preparedness, (β = −.16, Holm-adjusted p < .001). The strongest positive correlation with preparedness was with AI literacy (β = .24, all Holm adjusted p < .001); trusted-adult support (β = .19, all Holm adjusted p < .001); and reporting readiness (β = .31, all Holm adjusted p < .001). When these were added, marginal \(R^2\) went from .151 to .437.

Implications: The results indicate opportunities to assess the teaching and learning of AI-literacy, preparing trusted adults, the availability of confidential reporting avenues, and safeguarding protocols.

Conclusion: Preparedness was linked to individual skills and social supports and differed across schools. To establish temporality and causality, there is a need for longitudinal studies.