摘要
文章阐述生成式AI在开源情报工作中的应用价值,包括提升数据挖掘与获取效率、拓展数据处理与分析深度及提升情报传递直观性,分析面临的信息真实性、数据隐私、算法偏见等潜在风险,在此基础上提出开源情报工作中生成式AI应用风险应对策略,即构建开源情报数据源可靠性审查与生成内容准确性验证机制以应对信息真实性风险,构建数据分级保护机制、动态情报源监控机制和开源情报规模化处理保护机制以应对数据隐私风险,强化语料库审查与专项建设、构建多层次算法治理机制以应对算法偏见风险。
The article elaborates on the application value of generative AI in open source information work,including improving data mining and acquisition efficiency,expanding data processing and analysis depth,and enhancing the convey intuitiveness of information products,analyses potential risks faced in information authenticity,data privacy,and algorithm bias.Based on this,risk response strategies for the application of generative AI in open source information work are proposed,which includes building a reliablity review of open source information data source and accuracy verification mechanism of generated contents to address information authenticity risks,constructing a data classification protection mechanism,a dynamic intelligence source monitoring mechanism,and an open source information large-scale processing protection mechanism to address data privacy risks,strengthening corpus review and specialized construction,and establishing a multi-level algorithm governance mechanism to address algorithm bias risks.
作者
李逯炜
张梦星
Li Luwei;Zhang Mengxing
出处
《图书馆工作与研究》
北大核心
2025年第6期52-60,共9页
Library Work and Study
基金
中国人民公安大学基本科研业务费项目“情态证据侦查应用的科学原理和风险防控研究”(项目编号:2022JKF02052)研究成果之一。
关键词
国家安全
生成式AI
开源情报工作
风险
National security
Generative AI
Open source information work
Risk
作者简介
李逯炜(1994-),男,中国人民公安大学法学院2023级诉讼法学专业在读博士研究生,中国人民公安大学法学院,北京,100038;通讯作者:张梦星(1987-),女,副教授,中国人民公安大学侦查学院,北京,100038。
