Digital assistants: Inequalities and social context of access, use, and perceptual understanding

Abstract

This study focuses on digital divide in the context of access, use, and perceptual understanding of digital assistants. We pay particular attention to inequalities of perceptual outcomes that may be triggered by the first-(access) and second-level (use) divides. We extend this insight to the level of perceptual understanding and investigate how the understanding of various personalized AI-related applications—as manifested via the consumption of functional and informational features of digital assistants—vary depending on access and use. Our analyses of two U.S. national surveys reveal the first and second divides, such that those in higher status (higher income) enjoy higher access and use. Then, we also find related perceptual gaps along the line of socio-demographics, as the pattern was evident for education in interaction with other demographic backgrounds. That is, there were varying degrees of the understanding of algorithmic misjudgment, bias in recommended content, or data surveillance, while users with lower social status tended to easily overlook those risks for their excitement and convenience of AI-enabled devices. We argue that inequalities operate in terms of material (access/use) as well as socio-cultural (perceptual understanding) bases, suggesting how digital opportunities instigated by digital assistants may not be built on levelled grounds and continue to consolidate and reproduce existing inequalities in recursive ways.