生成式人工智能与图书馆读者隐私信息泄露的生成与防范——基于信息权力不对称理论视角  

Generative Artificial Intelligence and the Generation and Prevention of Leakage of Library Reader’Private Information:Based on Information Power Asymmetry Theory Perspective

作　　者：李玲 LI Ling

机构地区：[1]日照市图书馆,山东日照276800

出　　处：《图书情报导刊》2024年第11期7-13,共7页Journal of Library and Information Science

摘　　要：阐述了生成式人工智能环境下读者隐私泄露的严峻形势及其特征,分析了基于权力不对称理论视角下隐私泄露的形成机制,包括读者议价能力受限、图书馆信息优势、过度采集个人信息、信息不当使用等,提出了脱敏化处理规制信息采集权、合约化视角保障信息使用权、多中心治理视角保障信息管理权等防范读者隐私泄露的措施。This paper describes the severe situation and characteristics of reader privacy leakage in the generative AI environment,analyzes the formation mechanism of privacy leakage from the perspective of power asymmetry theory and proposes measures to prevent reader privacy leakage,including limited bargaining power of readers,information advan⁃tage of libraries,excessive collection of personal information,and improper use of information.To prevent privacy leak⁃age,the paper puts forward regulating information collection rights through desensitization processing,ensuring informa⁃tion usage rights from a contractual perspective,and guaranteeing information management rights from a multi-center gov⁃ernance perspective.

关 键 词：读者隐私 生成式人工智能 ChatGPT 信息权力不对称 

分 类 号：G252.0[文化科学—图书馆学]