摘要
生成式人工智能(GenAI)的迅猛发展推动了翻译数据的海量生成与高速传播,这些数据与国家信息安全紧密关联,其安全问题已成为维护国家总体安全的重要组成部分。本文首先界定GenAI翻译数据安全的内涵,提出其具有系统性、层级性和周期性特征,明确其构成要素为文本数据安全、元信息数据安全和用户数据安全。随后,本文分析GenAI时代翻译数据的风险表征,如数据滥用、数据篡改、隐私泄露和数据偏见等,并提出针对性的治理实践路径,如加强安全技术保障、完善数据管理机制、提升数据安全意识等,以期为国家翻译事业的数据安全保障提供参考。
The rapid advancement of Generative Artificial Intelligence(GenAI)has precipitated the massive generation and rapid dissemination of translation data,which is intrinsically linked to national information security.Consequently,the security of such data has emerged as a critical component in safeguarding holistic national security.This paper first delineates the conceptual connotations of GenAI translation data security,proposing that it exhibits systematicity,hierarchy,and cyclicity as its core characteristics.The components are identified as the security of textual data,metadata,and user data.Subsequently,the paper analyzes the risk manifestations characterizing the GenAI era,encompassing data abuse,data tampering,privacy breaches,and data bias.Correspondingly,targeted governance approaches are proposed,including strengthening security technology safeguards,improving data management mechanisms,and enhancing data security awareness,in order to provide insights for data security guarantees in the national translation endeavor.
作者
王贇
王均松
WANG Yun;WANG Junsong(Department of Foreign Languages,Taiyuan University,Taiyuan 030032,China;Confucius Institute,University College Dublin,Dublin D04V1W8,Ireland;School of Foreign Studies,Northwestern Polytechnical University,Xi’an 710061,China)
出处
《浙江外国语学院学报》
2026年第1期103-112,共10页
Journal of Zhejiang International Studies University
基金
教育部人文社会科学研究一般项目“基于受控语言的机器翻译自动译前编辑研究”(23YJAZH139)
教育部人文社会科学研究一般项目“中国古代科技典籍英译数据库构建与传播研究”(22YJC740081)
山西省教育科学“十四五”规划课题“‘AIGC+教育’理念下的翻译人才培养模式构建与应用研究”(GH-240707)。
