When the self meets the machine: Self-Schemas, symbolic fit, and engagement in AI–human service systems

This research investigates how consumers' self-perceptions, specifically self-objectification and self-assurance, shape their engagement with AI versus human service agents across appearance- and competence-based service encounters, and identifies mechanisms that can repair psychological mismatches in hybrid service ecosystems. Across four controlled experiments, participants engaged in scenario-based interactions that manipulated self-perception, service context, agent type, prosumption, and cognitive flexibility. Study 1 demonstrates a robust three-way interaction showing that engagement is highest when the agent–context pairing aligns with consumers’ underlying self-schema, mediated by psychological safety and relational warmth. Study 2 shows that introducing prosumption enhances engagement in mismatched human-delivered service encounters, consistent with the view that participatory service designs can alleviate engagement deficits arising from symbolic misfit. Studies 3A and 3B reveal that both induced and trait-level cognitive flexibility repair engagement deficits in mismatched AI encounters by enabling consumers to reinterpret the symbolic meaning of the interaction. Across all studies, the alignment between self-perception and the symbolic cues of the service encounter, rather than the absolute superiority of AI or human agents, emerges as the central determinant of consumer engagement. Together, the findings advance a consumer-centric, psychologically grounded model of AI–human service design and identify participatory and cognitive interventions that enable more inclusive and adaptive service experiences.