To be mild or be severe? Digital natives’ algorithmic resistance behavior in the mobile social media intelligent recommendation environment

Purpose In the context of mobile social media, users have adopted resistance practices to escape algorithmic deception and control, given that intelligent recommendation algorithms unintentionally infringe upon users’ autonomy while providing personalized services. This study aims to understand the antecedents and consequences of user algorithmic resistance behavior by examining its formation and influencing mechanisms within the mobile social media intelligent recommendation environment. Design/methodology/approach Semi-structured interviews and grounded theory were employed to investigate 32 digital natives with extensive experience in mobile social media. Through a three-level coding process, a theoretical model was constructed to elucidate the formation mechanism of user algorithmic resistance behavior. Findings Digital natives’ algorithmic resistance behavior, which can be classified into mild and severe categories, is driven by external stimuli (i.e. technical characteristics, social influence, and information characteristics) and internal stimuli (i.e. trust, affective state, privacy concerns, capacity, and risk assessment). Specifically, mild algorithmic resistance behavior results from rational resistance intention, while severe algorithmic resistance behavior is affected by irrational resistance intention. A bidirectional relationship is identified between behavioral performance and algorithm resistance behavior, indicating a mutual effect between mild and severe resistance behaviors. Behavioral performance could trigger the secondary or multiple resistance behaviors in terms of trust and affective state. Originality/value A theoretical model is constructed to explain the formation and influencing mechanisms of digital natives' algorithm resistance behavior within the mobile social media intelligent recommendation environment. It distinguishes between two specific types of algorithmic resistance, namely mild resistance and severe resistance, and identifies their unique transformation mechanisms. Additionally, this study elucidates the relationships among various influencing factors and provides a comprehensive and definitive characterization of the behavioral logic that emerges following the implementation of algorithm resistance behavior.