Grief is an inherent human experience that triggers distinctive coping responses. Advances in artificial intelligence now allow “thanabot” services, digital avatars that emulate deceased individuals, to offer novel forms of bereavement support. This study aims to investigate the psychological factors that drive the acceptance and use of these services.
Guided by Terror Management Theory (TMT), the research tests a theoretical model of thanabot acceptance through partial least squares structural equation modeling (PLS-SEM) and necessary condition analysis (NCA). An exploratory netnographic and documentary content analysis of online user narratives supplements and contextualizes the quantitative results.
Quantitative results show that escapism, coping and mortality salience act as mediators of thanabot acceptance. Realism interacts with social support to enhance perceived emotional benefits, especially during early engagement. Furthermore, lower levels of psychological readiness are identified as necessary preconditions for acceptance. The qualitative insights reinforce these mechanisms, revealing four emergent themes: symbolic continuity, emotional scaffolding, perceived realism and personalization and ethical ambivalence. Social support emerged as a recurring factor shaping emotional meaning and user engagement.
This research pioneers the application of TMT to AI-mediated grief technologies and proposes an integrated psychological framework for technology acceptance in emotionally sensitive contexts. The combined use of NCA and PLS-SEM advances service-research methodology by revealing both predictive pathways and boundary conditions, while qualitative insights enrich theoretical depth and inform the ethical design of thanabot services.
