- Daydreaming styles and self-compassion jointly predict adolescents' disclosure to AI chatbots.
- Adolescents with ADHD show distinct daydreaming and self-compassion profiles; self-compassion predicts reduced AI disclosure for them.
- AI 'Step 0' disclosures should be interpreted within stepped-care models and tailored for neurodiversity.
Br J Clin Psychol. 2026 Aug 17. doi: 10.1111/bjc.70087. Online ahead of print.
ABSTRACT
OBJECTIVE: AI chatbots are increasingly becoming a first common point of mental health disclosure for adolescents, suggesting a potential ‘Step 0′ within stepped-care pathways prior access to professional help. This study examined how daydreaming styles and self-compassion relate to adolescents’ disclosure of inner thoughts to an AI ‘listener’, and whether these psychological processes differ among young people with ADHD, for whom spontaneous mind wandering and attentional control difficulties may be especially salient.
DESIGN: A large-scale cross-sectional online survey was conducted with adolescents and young people from the United States and Hong Kong.
METHODS: Participants were 2115 adolescents with and without ADHD identification. Structural equation modelling was used to test hypothesized pathways between daydreaming styles, self-compassion and AI disclosure with multi-group analyses (ADHD vs. peers).
RESULTS: Youth with ADHD differed markedly from their peers in daydreaming styles and self-compassion. Pathways from daydreaming styles to AI disclosure also differed by ADHD status. Self-compassion showed context-dependent associations. Among youth with ADHD, higher self-compassion predicted reduced AI disclosure. In the general population group, self-compassion amplified disclosure associated with positive-constructive daydreaming, while constraining disclosure linked to guilty-dysphoric daydreaming and poor attentional-control daydreaming.
CONCLUSIONS: Self-compassion and daydreaming styles jointly influence adolescents’ tendencies to disclose mental health-related experiences to AI chatbots, with substantially different patterns among youth with ADHD. These findings suggest that AI-based ‘Step 0’ disclosures should be interpreted within stepped-care models with attention to individual differences and neurodiversity. Tailored, neurodiversity-informed approaches may be needed when integrating AI disclosure into youth mental health support pathways.
PMID:42606151 | DOI:10.1111/bjc.70087
Share Evidence Blueprint
Save to Google Notes
Search Google Scholar
Save as PDF
⭐ My Revision List
