Intelligence and humanness as key drivers of service value in Generative AI chatbots

Automated service agents with different levels of artificial intelligence (AI) - mechanical, analytical, and emotional - are gradually replacing employees in their interactions with customers. Previous research suggests that these agents (e.g., chatbots) should embed human behavioral traits such as warmth, competence, and even automated social presence (i.e., perceiving that one is interacting with someone else). However, it is unknown whether different levels of AI are perceived by customers as features increasing humanness and, subsequently, leading to higher functional, monetary, social, or emotional value. Through an experimental design based on tourism chatbots, the results from structural equation analysis reveal that whereas mechanical AI decreases automated social presence, analytical AI increases perceptions of competence and warmth, and emotional AI improves all humanness cues. The article merges the research streams of service agent design and customers’ perceptions of humanness to guide tourism managers in implementing generative AI agents to increase service value. • A lab study experimental design with a sample of 308 participants is conducted. • Mechanical AI reduces the perceptions of humanness in chatbots. • Analytical and emotional AI favors the perception of generative AI humanness. • Competence is associated with greater perceived value, including monetary value. • Warmth is associated with greater functional, social and emotional value.