Who has your back? Countering dark patterns in online shopping using interpersonal and AI-delivered support

Abstract

Online shopping interfaces often employ dark patterns to influence user behavior, leading to impulsive buying decisions. This study aims to enhance consumer protection by exploring how interventions incorporating various support sources (interpersonal, AI-delivered, or self) with message types (cognitive vs. affective) can mitigate the impact of dark patterns on impulsive buying behavior. Grounded in the Stimulus-Organism-Response (S-O-R) framework, this study theorizes intervention designs through a 3 × 2 between-subjects experiment (n = 363), examining how different sources and formats of support influence user responses to manipulative design. Mediators like emotional state, argument quality, and image appeal were included in the model. The findings indicate that AI-delivered support can reduce impulsive buying intentions, particularly when persuasive content and appealing visuals are integrated, highlighting the potential of well-designed machine-mediated interventions. Practically, the findings inform the development of AI-driven interventions that can be embedded into shopping platforms to promote more ethical consumer experiences. This research advances theory by demonstrating that AI outperforms interpersonal supports in reducing shopping impulses, offering insights for interface design for addressing dark pattern influences.