北京第二外国语学院学报 ›› 2026, Vol. 48 ›› Issue (3): 26-37.DOI: 10.12002/j.bisu.646

• 翻译研究(AI风险治理专栏 主持人:任东升) • 上一篇    下一篇

GenAI时代翻译技术伦理风险及其治理路径研究

莫爱屏1,2(), 张雅雯2(), 段胜峰3()   

  1. 1 广州商学院外国语学院广州 511363
    2 广东外语外贸大学高级翻译学院广州 510400
    3 长沙理工大学外国语学院长沙 410076
  • 出版日期:2026-06-30 发布日期:2026-07-08
  • 通讯作者: 张雅雯
  • 作者简介:莫爱屏,广州商学院外国语学院/广东外语外贸大学高级翻译学院教授,博士生导师,511363/510400,研究方向:翻译学、语言学及应用语言学。电子邮箱:moaiping@gdufs.edu.cn
    张雅雯,广东外语外贸大学高级翻译学院,510400,研究方向:翻译学、语言学及应用语言学。电子邮箱:20231220010@mail.gdufs.edu.cn
    段胜峰,长沙理工大学外国语学院教授,410076,研究方向:翻译学、跨文化交际。电子邮箱:duansf1077@163.com
  • 基金资助:
    国家社会科学基金项目“中华文化外译的语用策略研究”(17BYY056);广东省哲学社会科学规划青年项目“人机竞译——MTI学生与AI翻译质量对比研究”(GD24YWY07);广东外语外贸大学研究生科研创新项目“中国特色文化话语英译的多模态语用策略研究”(24GWCXXM-083)

Ethical Risks of Translation Technology and Governance Pathways in the GenAI Era

Mo Aiping1,2(), Zhang Yawen2(), Duan Shengfeng3()   

  1. 1 Guangzhou College of Commerce, Guangzhou 511363, China
    2 Guangdong University of Foreign Studies, Guangzhou 510400, China
    3 Changsha University of Science and Technology, Changsha 410076, China
  • Published:2026-06-30 Online:2026-07-08

摘要:

生成式人工智能(GenAI)在当今人文社会科学领域引发广泛关注。基于对伦理学、技术哲学及技术伦理学等领域理论的思考,本研究构建了“人 - 数据 - 社会”三维分析框架,系统探讨翻译技术伦理在GenAI语境下的伦理风险:①人机关系维度的译者身份漂移与版权归属模糊;②数据监管维度的隐私泄露、数据失真与算法偏见;③社会影响维度的资源垄断与语言同质化。针对上述问题,本研究倡导“人机协同 - 数据共治 - 伦理共建”的三重治理路径,以期为翻译技术伦理的规范和可持续创新提供可操作的政策建议与学理支撑。

关键词: 生成式人工智能; 翻译技术; 技术伦理; 翻译技术伦理; 翻译学

Abstract:

Generative artificial intelligence (GenAI) has attracted widespread attention across the humanities and social sciences. Drawing on insights from ethics, the philosophy of technology, and technology ethics, this study constructs a three-dimensional analytical framework of “human-data-society” to systematically examine the ethical risks of translation technologies in the context of GenAI. At the human-machine relationship level, the risks involve translator identity shift and ambiguous copyright ownership; at the data governance level, the risks include privacy breaches, information distortion, and algorithmic bias; and at the societal impact level, the risks manifest as resource monopolization and linguistic homogenization. In response, this study advocates a threefold governance path of “human-machine collaboration, data co-governance, and joint ethical construction”, aiming to provide actionable policy recommendations and theoretical support for the regulation and sustainable innovation of translation technology ethics.

Keywords: GenAI; translation technology; technology ethics; translation technology ethics; translation studies

中图分类号: