华东师范大学学报(教育科学版) ›› 2026, Vol. 44 ›› Issue (6): 92-102.doi: 10.16382/j.cnki.1000-5560.2026.06.006

• 教育智能化转型 • 上一篇    下一篇

教育人工智能伦理风险评估指标构建与应用研究

王佑镁1,2, 黄芳琴2, 柳晨晨2   

  1. 1. 浙江东方职业技术学院,温州 325000
    2. 温州大学大数据与智慧教育研究中心,温州 325035
  • 出版日期:2026-06-01 发布日期:2026-05-27
  • 基金资助:
    国家社会科学基金教育学重大课题“新一代人工智能对教育的影响研究”(VGA230012);认知智能全国重点实验室智能教育开放课题“新一代人工智能在教育领域的伦理风险防范” (iED2024-Z001)。

Research on the Construction and Application of Ethical Risk Assessment Indicators for Artificial Intelligence in Education

Youmei Wang1,2, Fangqin Huang2, Chenchen Liu2   

  1. 1. Research Center for Artificial Intelligence in Vocational Education, Zhejiang Dongfang Polytechnic, Wenzhou Zhejiang 325000, China
    2. Research Center for Big Data and Smart Education, Wenzhou University, Wenzhou Zhejiang 325035, China
  • Online:2026-06-01 Published:2026-05-27

摘要:

作为引领新一轮科技革命的战略性技术,人工智能技术在深刻改变教育教学生态的同时,也带来了诸多伦理风险及挑战。如何评估伦理风险以最大化享受智能技术红利,成为教育人工智能健康发展的重要议题。本研究通过理论研究、专家访谈、文献调研等方式,确定了教育技术本体风险、教育数据风险、机器算法风险、教育应用风险4个一级指标以及13个二级指标和31个三级指标,运用层次分析法(AHP)设计相应问卷,利用MATLAB软件对各指标权重进行模拟仿真计算,形成系统性较强的伦理风险评估指标体系,并选取案例进行评估应用确证其有效性。构建教育人工智能应用的伦理风险评估指标体系,可用于在具体教育场景中识别和评估风险,助力教育管理部门、学校和师生规范合理应用人工智能技术改进教学,并为教育人工智能治理提供依据。

关键词: 教育人工智能, 伦理风险, 层次分析法, 评估指标

Abstract:

As a strategic technology leading the new round of scientific and technological revolution, Artificial Intelligence (AI) technology, while profoundly changing the ecology of education and teaching, also brings many ethical risks and challenges. How to assess ethical risks in order to maximize the benefits of intelligent technology has become an important issue for the healthy development of educational artificial intelligence. Through theoretical research, expert interviews, literature research, etc., four primary indicators, 13 secondary indicators, and 31 tertiary indicators are identified, such as the risk of educational technology ontology, the risk of educational data, the risk of machine algorithms, and the risk of educational applications. The corresponding questionnaires are designed using the Analytic Hierarchy Process, and the weights of the indicators are calculated using MATLAB software to establish a systematic and robust ethical risk assessment system. Cases are then selected for evaluation to validate its effectiveness. The construction of the ethical risk assessment index system applicable to the application of AI in education can be used to identify and assess the risks in specific educational scenarios, help education management departments, schools, teachers, and students standardize and rationally apply AI technology to improve teaching and learning, and provide a basis for the governance of AI in education.

Key words: artificial intelligence in education, ethical risk, Analytic Hierarchy Process, assessment indicators