The Influence of ‘Algorithm Aversion’ and ‘Algorithm Appreciation’ Among Consumers in Unstructured Tasks

ABSTRACT Consumer algorithmic trust represents a research topic of both theoretical and practical significance. Existing research has identified two key phenomena within the domain of consumer algorithmic trust: algorithm aversion and algorithm appreciation. However, the majority of prior studies have primarily focused on one of these two phenomena. The present research demonstrates that consumers exhibit both algorithm aversion and algorithm appreciation across various unstructured tasks. Utilizing a text analysis from Douyin (the Chinese TikTok) and three experimental studies, it is demonstrated that consumers have higher trust in algorithm agents compared with human agents in lower‐degree unstructured tasks, with perceived accuracy mediating this effect. Conversely, higher trust in human agents compared with algorithm agents is observed in higher‐degree unstructured tasks, with perceived flexibility mediating this effect. This study contributes to the extant literature on consumer algorithmic trust and offers actionable insights for the development of artificial intelligence systems.