Investigating how AI Personalization Algorithms Influence Self-Perception, Group Identity, and Social Interactions Online

This study investigates the impact of AI-driven personalization on self-perception, group identity, and social interactions across major social media platforms, including Instagram, Twitter, and YouTube. Utilizing a quantitative approach, data was collected from 500 participants using stratified random sampling to ensure a diverse representation across age, gender, education, and geographic location. The study tests three hypotheses: (H1) AI personalization negatively affects self-perception, increasing anxiety and lowering self-esteem; (H2) personalized content strengthens group identity, leading to greater polarization; (H3) AI reduces social interactions by limiting exposure to diverse viewpoints. Descriptive statistics revealed moderate engagement with personalized content (M = 3.48, SD = 1.12) and notable anxiety (M = 3.26, SD = 1.23). Correlation analysis showed significant negative relationships between personalized content and self-esteem (r = -0.45) and anxiety (r = -0.53), while positive correlations were found between personalized content and in-group bias (r = 0.62). Regression analysis confirmed that personalized content exposure predicted lower self-esteem (? = -0.45), stronger in-group bias (? = 0.62), and reduced interaction with out-groups (? = -0.48). ANOVA results indicated significant differences in anxiety and self-esteem across age groups, with younger participants exhibiting higher anxiety and lower self-esteem. These findings support the hypotheses, showing that AI personalization can negatively affect self-perception, foster group identity, and limit social interactions by curating content that aligns with pre-existing preferences.