Unraveling the mechanisms of AI system aversion among customer-contact employees: a perspective from advice response theory

Abstract

Purpose

Based on the advice response theory perspective, this study aims to investigate the effects of human managers and artificial intelligence (AI) systems on customer-contact employees’ aversion to AI systems in the hospitality industry. It examined the mediating role of advice content characteristics (efficacy, feasibility and implementation limitations) and advice delivery (facework and comprehensibility) on customer-contact employees’ aversion to AI systems.

Design/methodology/approach

Two scenario-based experiments were conducted (Nexperiment 1 = 499 and Nexperiment 2 = 300). Experiment 1 compared the effects of different advisor types (human managers vs AI systems) on employees’ aversion to AI systems. Experiment 2 investigated the mediating role of advice content characteristics (efficacy, feasibility and implementation limitations) and advice delivery (facework and comprehensibility).

Findings

The results showed employees tended to prioritize advice from human managers over output from AI systems. Moreover, advice content characteristics (efficacy, feasibility and implementation limitations) and advice delivery (facework and comprehensibility) played mediating roles in the relationship between advisor type characteristics and employees’ aversion to AI systems.

Practical implications

These findings contribute to the understanding of AI system aversion and provide theoretical insights into management practices involving customer-contact employees who interact with AI technology in the hospitality industry.

Originality/value

The primary contribution of this study is that it enriches the literature on employee aversion to AI systems by exploring the dual mediators (advice content characteristics and advice delivery) through which advisor type characteristics affect AI system aversion.