•  
北京第二外国语学院学报

北京第二外国语学院学报 ›› 2026, Vol. 48 ›› Issue (3): 38-50.DOI: 10.12002/j.bisu.647

• 翻译研究（AI风险治理专栏 主持人：任东升） • 上一篇    下一篇

生成式AI时代的翻译安全风险及其治理研究

高玉霞(), 任东升()   

  1. 中国海洋大学， 青岛 266100
  • 出版日期:2026-06-30 发布日期:2026-07-08
  • 作者简介:高玉霞，博士，中国海洋大学副教授，266100，研究方向：国家翻译安全研究。电子邮箱：gaoyuxia@ouc.edu.cn;
    任东升，博士，中国海洋大学教授，266100，研究方向：国家翻译研究。电子邮箱：dongsheng_ren@ouc.edu.cn
  • 基金资助:
    教育部人文社会科学基金青年项目“美智库涉华翻译的实践、政策及中国应对研究（2017—2024）”(24YJC740018);2025年度山东省自然科学基金青年项目“美智库对华生物安全开源情报翻译及其转化应用对我国生物安全的影响评估”(ZR2025QC808);2024年度山东省博士后创新项目“中美博弈视域下美国智库涉华科技开源情报翻译研究”(SDCX-RS-202400022)

Translation Security Risks and Their Governance in the Generative AI Era

Gao Yuxia(), Ren Dongsheng()   

  1. Ocean University of China, Qingdao 266100, China
  • Published:2026-06-30 Online:2026-07-08

RichHTML

PDF (PC)

摘要：

生成式AI凭借强大的机器学习和自然语言处理能力给翻译行业、翻译学科教育及其他领域带来了转型升级机遇，但同时也带来了诸多安全隐患。生成式AI时代的翻译安全风险具有多样态、内嵌性和泛在化等特点，对翻译安全治理提出了严峻挑战。本文在梳理生成式AI时代的翻译安全风险来源及特点的基础上，提出如下治理对策：应坚持在总体国家安全观指引下，均衡把握技术价值创造、敏感数据保护和国家安全维护等综合需求，采取技术、教育和制度融贯一体的治理模式。在这一思路指引下，应优先考虑采用AI、区块链等技术手段应对技术问题；着力加强科普和专业教育，提升全民生成式AI翻译安全意识和安全能力；还应考虑采取政策、立法等制度手段，控制治理成本，提高治理效能。

关键词: 生成式AI; 翻译安全; 风险; 治理; 总体国家安全观

Abstract:

The advent of generative AI, with its advanced machine learning and natural language processing capabilities, presents transformative opportunities for the translation service industry, translation education, and other sectors, while also introducing significant security risks. In the era of generative AI, translation security risks are characterized by multiplicity, embeddedness, and ubiquity, which pose great challenges to translation security governance. Based on an analysis of the sources and characteristics of translation security risks in the era of generative AI, this paper proposes the following governance strategies: adhering to a holistic view of national security, balancing comprehensive demands such as technological value creation, sensitive data protection, and national security maintenance, and adopting a governance model that integrates technology, education, and institutions. Priority should be given to technological solutions such as AI and blockchain to address technical problems, and efforts should also be made to strengthen science popularization and professional education to raise the public’s awareness and enhance their security capabilities in the use of generative AI translation. Furthermore, policy and legislative measures should be considered to control governance costs and improve governance effectiveness.

Keywords: generative AI; translation security; risks; governance; a holistic view of national security

中图分类号: