生成式人工智能用户虚假信息规避行为研究  

Research on False Information Avoidance Behavior of Generative Artificial Intelligence Users

作　　者：常琳 叶凤云[1] Chang Lin;Ye Fengyun(School of Management,Anhui University,Anhui Hefei 230601)

机构地区：[1]安徽大学管理学院,安徽合肥230601

出　　处：《情报理论与实践》2025年第2期102-112,共11页Information Studies:Theory & Application

基　　金：国家社会科学基金项目“认知与情绪双路径视角下社交网络虚假信息用户趋避行为研究”的成果,项目编号:23BTQ060。

摘　　要：[目的/意义]探究生成式人工智能情境中用户规避虚假信息的行为模式及其形成机制,揭示用户虚假信息规避行为过程中的认知、情感及行为变化,提高用户的信息素养和增强防范虚假信息的能力。[方法/过程]采用半结构化访谈形式对30名生成式人工智能用户进行调研,基于CAPS理论及利用扎根理论研究方法,对收集到的访谈数据进行编码分析,提炼出4个核心范畴和18个主范畴。[结果/结论]在生成式人工智能虚假信息情境中,用户会经历认知—情感—行为的动态演变过程。情境因素触发用户个性系统中的认知—情感单元,引发认知评估和情感反应,进而形成虚假信息规避行为;行为结果又会作为新的情境因素反馈给用户,影响其后续的认知评估和情感反应,形成循环动态过程,不断塑造和强化用户的虚假信息规避行为。[Purpose/significance]This study aims to explore the behavioral patterns and formation mechanisms of users’avoidance of false information in generative artificial intelligence,and to reveal the cognitive,emotional,and behavioral changes during the process of avoiding false information,in order to improve users’information literacy and enhance their ability to prevent false information.[Method/process]A semi-structured interview was conducted with 30 users of generative artificial intelligence.Based on the CAPS theoretical framework and using grounded theory research methods,the collected interview data were coded and analyzed,resulting in 4 core categories and 18 main categories.[Result/conclusion]Users undergo a dynamic evolution process of cognition-emotion-behavior when dealing with false information in generative artificial intelligence.Situational factors trigger cognitive-emotional units within the user's personal system,leading to cognitive assessment and emotional responses.These then lead to behaviors for avoiding false information.The results of user behavior will then be fed back as new situational factors affecting subsequent cognitive assessments and emotional responses forming a cyclical dynamic process that continuously shapes strengthens users'avoidance behaviors towards false information.

关 键 词：生成式人工智能 虚假信息 规避行为 CAPS理论 扎根理论 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程] G252[自动化与计算机技术—控制科学与工程]