This study aims to examine the influence of greedy and biased recommendation algorithms on consumers’ emotional arousal, impulsive buying behavior and post-purchase dissonance in artificial intelligence (AI)-driven short video platforms, such as TikTok, Instagram Reels and YouTube Shorts. It further investigates the moderating roles of promotion-proneness and inferences of manipulative intent.
A quantitative survey was conducted with 766 Vietnamese consumers who had recently made purchases influenced by AI recommendations. Partial least squares structural equation modeling was used to test a conceptual framework grounded in cognitive dissonance theory and behavioral decision theory, incorporating mediation and moderation analyses.
Greedy and biased algorithms significantly enhance emotional arousal, which mediates the pathway to impulsive buying. Impulsive buying, in turn, heightens post-purchase dissonance. Promotion-proneness strengthens the arousal to impulsive buying link, while inferences of manipulative intent amplify the impulsive buying to dissonance relationship. Serial mediation effects were fully supported, with no direct algorithmic impacts on dissonance.
This study relies on cross-sectional, self-reported data collected from Vietnamese consumers, which may limit causal interpretation and generalizability to other cultural and market contexts. Future research should use longitudinal, experimental or behavioral-tracking designs and examine other algorithmic characteristics, cultural settings and psychological factors influencing consumer responses to AI-driven recommendations.
The findings suggest that platform designers and marketers should balance engagement-oriented personalization with consumer autonomy and well-being. Greater transparency regarding algorithmically selected or commercially prioritized content, meaningful user-control mechanisms and less manipulative promotional practices may help reduce impulsive purchasing and post-purchase dissonance. Firms may also use post-purchase reassurance, customer support and satisfaction guarantees to reduce consumers’ regret and maintain long-term trust.
This research provides novel empirical evidence linking algorithmic design features to emotional and cognitive mechanisms in impulsive consumption. It advances understanding of AI’s ethical implications in digital commerce, highlighting psychological costs in emerging markets and informing strategies for balanced personalization.
