Conversational Artificial Intelligence (CAI) systems are increasingly used by consumers as service advisors, guides, and support for emotionally sensitive and care-oriented domains such as pet health, nutrition, behaviour, and welfare. As consumers increasingly rely on these AI-mediated service interactions, it has become an important service marketing concern to understand these systems and their role in shaping emotional experiences. While existing research has primarily evaluated CAI platforms on dimensions such as accuracy, transparency, and ethical risk, considerably less attention has been paid to the role of CAIs in shaping users’ emotional experiences, influence on emotional burden, emotional reassurance and their metaemotional responses—how users feel about their own emotions when interacting with Conversational AI.

To address this gap this study examines CAI as human-centred service advisor through a qualitative comparative analysis of five widely used CAI platforms - ChatGPT, Gemini, Perplexity, Copilot and Meta AI, focusing on their responses to identical pet care–related queries. A prompt-controlled research design, using a standardized set of pet care queries was developed to use across four categories: routine care, dietary decisions, health concerns, and ethically ambiguous situations. Responses from each CAI system were systematically analyzed using a two-layer coding framework capturing (1) primary emotional cues embedded in AI responses (e.g., reassurance, urgency, caution, empathy) and (2) metaemotional affordances, including emotional regulation, amplification of anxiety or confidence, guilt normalization, and dependence signalling. Cross-platform comparison was conducted to identify systematic differences in emotional framing, uncertainty communication, and reflective guidance.

Findings demonstrate that CAI platforms create distinct human-centred service experiences by varying in their emotional reassurance, ethical framing, communication of uncertainty and support for users’ confidence, anxiety, guilt, and responsibility. The study contributes to service marketing and human-AI interaction literature by introducing the META-CARE (Metaemotional and Ethical Care) Framework, a qualitative evaluation tool for emotionally sensitive and AI-mediated service experiences. Beyond technical performance, the framework highlights the role of  CAI in shaping emotional burden, metaemotional affects, consumer trust, and decision confidence. It offers practical guidance for designing more human-centred AI service interactions in care-oriented contexts, including pet care....