Online aggression in Czech adolescents: prevalence, demographic variation, and platform-related risk

This study examines the prevalence and patterns of online aggression among Czech adolescents, including emerging forms such as AI-generated sexual imagery, using a large cross-sectional school-based survey of 42,772 respondents aged 13–17 conducted in 2025. Data were collected via an anonymous online questionnaire capturing 13 behaviourally specific forms of online aggression, and analysed primarily using descriptive statistics, chi-square tests, effect sizes, and a respondent-fixed-effects logistic regression model for platform-specific risk. The results show that online aggression is highly prevalent, with verbal aggression reported by 58.6% of adolescents and 31.9% experiencing regular exposure; notable findings include the emergence of AI-generated nude imagery (7.3%), higher prevalence in vocational school tracks, and a robust association between platform use intensity and reported harm. These findings suggest that online aggression represents a widespread and multifaceted risk affecting adolescents, extending beyond traditional cyberbullying definitions, with important implications for prevention, policy, and the need to address AI-enabled forms of harm and low rates of disclosure among victims.