心理科学进展2026，Vol.34Issue(1)：123-133,11.DOI:10.3724/SP.J.1042.2026.0123
算法介导下的情感趋同:生成式人工智能情感传染机制
Algorithm-mediated emotional convergence:The emotional contagion mechanisms of artificial intelligence generated content
摘要
Abstract
This study addresses the emerging phenomenon of emotion contagion in Artificial Intelligence Generated Content(AIGC),systematically examining its fundamental distinctions from both traditional interpersonal emotion contagion and digital emotion contagion.It proposes the"enactment-modulation"mechanism as a theoretical framework.The research identifies four key characteristics of AIGC-driven emotion contagion—intersubjectivity,knowledge dependency,non-threatening and de-identified nature,and moral relevance—which collectively underpin the"enactment-modulation"mechanism.Specifically,AIGC enacts human-like emotional expression patterns through algorithmic simulation("enactment"),while dynamically refining interaction strategies based on real-time user feedback("modulation"),thereby forming a continuously evolving human−machine emotional feedback loop.The"enactment-modulation"mechanism transcends anthropocentric paradigms,contributes to cross-subjective emotion theory,and reveals the novel role of algorithms as proactive emotion modulators.It has already found applications in domains such as mental health intervention,communication studies,and educational motivation.Research on AIGC emotion contagion extends the theoretical scope of emotion contagion studies and offers a fresh perspective for understanding human−AI emotional interaction.However,it also confronts several challenges,including difficulties in multidisciplinary integration,complexities in multimodal emotion measurement,barriers to cross-cultural adaptation,and the risk of emotional misdirection due to algorithmic bias.关键词
生成式人工智能(AIGC)/情感传染/人机交互Key words
Artificial Intelligence Generated Content(AIGC)/emotional contagion/human-AI interaction分类
社会科学引用本文复制引用
武靖宇,金鑫..算法介导下的情感趋同:生成式人工智能情感传染机制[J].心理科学进展,2026,34(1):123-133,11.基金项目
省部级基金"多模态情感传播视域下信息失序风险识别与网络戾气治理研究"(24YJA86008)、重庆市教委一般项目"数字时代图像新闻的情感传播与影响机制研究"(24SKGH062)资助. （24YJA86008）
