Exploring the Relationship Between Fear of Being Alone, Psychological Distress, and Problematic ChatGPT Use: A Cross-Sectional Study Among Young Adults.

Artificial Intelligence (AI)-based conversational agents such as ChatGPT have become deeply embedded in daily lives of young adults; understanding the psychological factors associated with their problematic use has become an urgent area of inquiry. Drawing on the Compensatory Internet Use model (CIU; Kardefelt-Winther, 2014), this study examined whether fear of being alone and psychological distress (depression, anxiety, and stress) were associated with problematic ChatGPT use among national and international young adults aged 18-35. While depression and anxiety have been linked to problematic ChatGPT use in prior research, fear of being alone despite sharing conceptual features with established predictors such as avoidance-based coping and unmet need for connection- remains unexamined in this context. A correlational and cross-sectional design was employed, with data collected through an anonymous online survey from a sample of national and international young adults (N = 50, Mage = 24.10, SD = 3.87). Participants completed the Brief Scale of Fear of Loneliness (BSFL), the Depression Anxiety Stress Scales-21 (DASS-21), and the Problematic ChatGPT Use Scale (PCUS). Spearman’s rank order correlations indicated that psychological distress was significantly and positively associated with problematic ChatGPT use (ρ = .35, p < .05), with depression (ρ = .43, p < .01) and anxiety (ρ = .29, p < .05) emerging as the strongest contributors within the framework, while stress did not reach significance. Fear of being alone showed a positive trend in the hypothesized direction (ρ = .25, p > .05) but did not attain statistical significance. Mann-Whitney U tests indicated no significant differences between national and international participants on any study variable. These findings are broadly consistent with the CIU model suggesting that internalized emotional difficulties such as depression and anxiety - rather than general situational stress or fear of being alone — are more closely linked to problematic ChatGPT use. This study contributes to the emerging literature by being, to our knowledge, the first to empirically examine fear of being alone in relation to problematic ChatGPT use, and by extending the CIU framework to a novel AI context. These findings carry implications for young adults’ well-being services and mental health practitioners working therapeutically with this population.