Speciesism toward AI: The mechanism of AI affection expression on user satisfaction

Abstract

Evaluations of AI services may be adversely influenced by speciesism due to AI's distinctiveness from humans. We conducted four studies to investigate the impact of speciesism on AI evaluations. Our findings suggest that AI's capacity for affection expression (AE) increases people's satisfaction with AI services by reducing their speciesism toward AI. Additionally, we found that the influence of speciesism on AI satisfaction is moderated by the type of task. This study offers novel recommendations for AI service providers and designers.