Understanding users' negative responses to recommendation algorithms in short-video platforms: a perspective based on the Stressor-Strain-Outcome (SSO) framework

AI-based recommendation algorithms have received extensive attention from both academia and industry due to their rapid development and broad application. However, not much is known regarding the dark side, especially users' negative responses. From the perspective of recommendation features and information characteristics, this study aims to uncover users' negative responses to such AI-based recommendation algorithms in the algorithm-driven context of short-video platforms. Drawing on the stressor-strain-outcome (SSO) framework, this study identifies information-related stressors and examines their influence on users' negative responses to a recommendation algorithm. The results show that such algorithms' greedy recommendation feature induces information narrowing, information redundancy, and information overload. These information factors predict users' exhaustion, which in turn promotes users' psychological reactance and discontinuance intention. This study adds knowledge on the dark side of recommendation algorithms.