Impact of artificial intelligence-generated self-images on children's body image development: a cross-sectional study in Mexico

Generative artificial intelligence (G-AI) has transformed children’s digital environments through tools that alter self-image using idealized filters and avatars. Early exposure to these technologies poses psychological risks, particularly regarding body perception during sensitive developmental stages. Objective To analyze the psychological effects of exposure to Generative Artificial Intelligence self-images on the body image development of Mexican children aged 6 to 12 years. Methods A quantitative, cross-sectional, and correlational study was conducted with a non-probabilistic sample of 302 children. Standardized scales were used to assess body satisfaction, G-AI exposure, and parental mediation. Descriptive statistics, ANOVA, principal component analysis (PCA), cluster analysis, and multiple linear regression were performed. Results A significant negative correlation was found between G-AI use and body satisfaction (r = -0.42, p < .001), while parental mediation showed a positive association (r = 0.31, p < .01). Multiple regression analysis confirmed both G-AI exposure and parental mediation as significant predictors of body satisfaction (adjusted R2 = 0.18). Three user profiles were identified based on exposure level, body esteem, and adult supervision. Conclusion Frequent use of G-AI can negatively impact children's body image, especially without parental mediation, highlighting the need for preventive educational and family-based strategies. Conceptual model illustrating the relationship between exposure to generative AI, parental mediation, and body image satisfaction in children. Arrows represent hypothesized directions based on theoretical models (Bandura 1986; Vygotsky 1978).