信息资源管理学报 ›› 2026, Vol. 16 ›› Issue (1): 37-49.doi: 10.13365/j.jirm.2026.01.037

• 学术专栏-网络暴力与虚假信息治理 • 上一篇    下一篇

基于人工智能生成内容的深度伪造事件舆情信息传播干预机制研究

杨洋洋   

  1. 郑州轻工业大学经济与管理学院,郑州,450001
  • 出版日期:2026-01-26 发布日期:2026-03-23
  • 作者简介:杨洋洋,博士,副教授,研究方向为网络舆情与信息治理, Email:2022016@zzuli.edu.cn。
  • 基金资助:
    本文系国家社会科学基金青年项目“ 基于人工智能生成内容的深度伪造事件舆情信息风险感知与场景治理研究”(24CTQ042)的研究成果之一。

Research on the Intervention Mechanism of Public Opinion Information Dissemination in Deepfake Events Based on AIGC

Yang Yangyang   

  1. School of Economics & Management, Zhengzhou University of Light Industry, Zhengzhou, 450001
  • Online:2026-01-26 Published:2026-03-23
  • About author:Yang Yangyang, Ph.D., associate professor, research interests including online public opinion and information governance, Email: 2022016@zzuli.edu.cn.
  • Supported by:
    This work is one of the research results of the National Social Science Fund Youth Project "Research on Public Opinion Information Risk Perception and Scenario Governance of Deepfake Events Based on AI-Generated Content" (24CTQ042).

摘要: 基于信任理论,结合人工智能生成深度伪造事件舆情信息的特征,从信息策略、责任策略和行为策略三类传播干预策略入手展开分析,构建了“干预策略-感知信任-干预后信任”的研究框架,探讨干预策略对系统信任、人际信任、情感信任、认知信任的影响机制。研究发现,信息策略通过人际信任和认知信任间接正向影响干预后信任;责任策略通过系统信任和情感信任间接正向影响干预后信任;行为策略通过系统信任和人际信任间接正向影响干预后信任;控制变量(性别、年龄和教育)对干预后信任的影响不显著;不同性别类型下干预策略对感知信任的影响不显著。

关键词: 深度伪造事件, 舆情信息, 干预策略, 信任理论, 中介模型

Abstract: Based on trust theory and the characteristics of deepfake event public opinion information generated by artificial intelligence, this paper analyzes the impact mechanism of intervention strategies on system trust, interpersonal trust, emotional trust, and cognitive trust from three types of communication intervention strategies: information strategy, responsibility strategy, and behavior strategy, with the research framework of "intervention strategy-perceived trust-post-intervention trust". The study finds that information strategy indirectly and positively affects post-intervention trust through interpersonal trust and cognitive trust, responsibility strategy indirectly and positively affects post-intervention trust through system trust and emotional trust, and behavioral strategy indirectly and positively affects post-intervention trust through system trust and interpersonal trust, while control variables (gender, age, and education) have no significant impact on post-intervention trust. The influence of intervention strategies on perceived trust does not vary significantly across different gender types.

Key words: Deepfake incidents, Public opinion information, Intervention strategies, Trust theory, Mediation model

中图分类号: