Honest machines? A cross-disciplinary perspective on trustworthy technology for children

Abstract

Humans increasingly interact with social robots and artificial intelligence (AI) powered digital assistants in their daily lives. These machines are usually designed to evoke attributions of social agency and trustworthiness in the human user. Growing research on human-machine-interactions (HMI) shows that young children are highly susceptible to design features suggesting human-like social agency and experience. Older children and adults, in contrast, are less likely to over attribute agency and experience to machines. At the same time, they tend to over-trust machines as informants more than younger children. Based on these findings, we argue that research directly comparing the effects of HMI design features on different age groups, including infants and young children is urgently needed. We call for evidence-based evaluation of HMI design and for consideration of the specific needs and susceptibilities of children when interacting with social robots and AI-based technology.