Key points are not available for this paper at this time.
Many young adults now turn to generative AI chatbots for emotional support, yet little is known about how they actively shape these interactions and regulate their psychological boundaries. We conducted semi-structured interviews with 22 Chinese young adults ages 18 to 32 who used generative artificial intelligence (AI) chatbots for stress or mental health-related concerns. Reflexive thematic analysis showed that use was driven by expressive motives, including venting, emotional labeling, and immediate relief, and relational motives, including companionship, validation, and freedom from burdening others. Participants described internalizing the chatbot's calm and structured dialogue to regulate emotions. At the same time, they maintained cognitive distance and enacted boundaries through selective disclosure, persona framing, iterative prompt refinement, and deliberate disengagement. We conceptualize this pattern as bounded supportive companionship, an emotionally meaningful yet intentionally limited form of AI-mediated support that reframes reciprocity as responsive turn-taking without enduring relational obligation.
Ma et al. (Sat,) studied this question.