Perceived coping and healing among Chinese emerging-adult users: The shaping role of AI emotional companion agents.

AI emotional companion agents are increasingly used for everyday emotional support, yet evidence remains limited regarding how observable service cues are translated into proximal, nonclinical support value during an initial encounter. Drawing on a stimulus-organism-response (SOR)-based Service-Psychology-Value (SPV) framework, this study examined 300 Chinese emerging adults aged 18-25 after a standardized first-use experience with an AI animal-companion application. Partial least squares structural equation modeling (PLS-SEM) supported the proposed direct and indirect associations: sense of companionship and affinity were associated with early relational closeness; relational closeness was associated with security and privacy and context-specific emotional-management efficacy; and all three psychological appraisals were associated with perceived coping and healing. Perceived coping and healing was operationalized as an immediate service-value appraisal combining perceived coping enablement and felt restoration, rather than as a measure of the complete coping process or clinical healing. A supplementary artificial neural network (ANN) analysis reproduced the same within-sample importance ordering, with intimacy ranked first and self-efficacy and security/privacy closely following. The ANN was retained as an alternative functional-form robustness check rather than as evidence of nonlinear causality or external predictive superiority. Follow-up interviews with 20 users identified three convergent themes: accessible emotional entry, relational interpretation, and immediate coping-oriented relief. Theoretically, SPV adds explanatory value by specifying a relational-conversion logic through which service cues acquire value only when they are interpreted as safety, closeness, and usable emotional agency. Practically, initial-session design should prioritize responsive presence, transparent data control, and low-pressure emotional tools. The findings apply to immediate perceptions following bounded exposure to nonclinical conversational companion services with similar configurations.