教育领域人工智能伦理研究进展:基于活动理论的分析框架

Research Advances on Ethical Artificial Intelligence in Education:an Analytical Framework based on Activity Theory

随着人工智能在教育领域应用的不断深化,教育伦理风险问题随之产生,引起了学术界的高度关注。采用活动理论分析框架,对2012—2022年间有关教育领域人工智能应用伦理风险的150篇文献进行系统性回顾,发现教育领域人工智能伦理研究经历了萌芽探索、成长发生、快速发展三个阶段;研究议题以安全隐私为主,但研究主体尚不明晰,对引起教育领域伦理问题的人工智能技术缺乏明确指向性;研究方法以定性为主,引发伦理问题的应用场景相对局限,且较少针对特定教育群体;研究对伦理困境关注较多,但风险管理视角欠缺;研究成果多为单一视角的理论与现状分析,缺少多理论视域的综合研究;伦理风险的跨领域交叉研究和多视域理论分析,对当下教育伦理及风险治理具有重要意义。基于主要结果及局限性,提出未来研究建议:需要重新审视人工智能与教育之间的交互作用,重塑人工智能教育应用中的伦理规约与评价指标;研究方法不局限于理论研究,应在人工智能技术实践教学中客观收集教学案例、资料等数据,科学性和创新性地提出教育领域人工智能伦理治理策略;深入探讨不同教育层级和教育领域的具体问题,进而针对性地制定相关政策和法规,加强监管与管理,并且提高相关人员的伦理意识和责任意识;不仅要关注教育人工智能应用表层所涉及的伦理问题,对于更深层次的风险发生机制问题也要多加关注;加大对教育人工智能应用伦理原则与政策的研究力度,系统性地思考人工智能的治理维度,通过制定伦理准则,加强政府及相关管理部门对教育人工智能的监管。

With the deepening of application of artificial intelligence in the field of education,the issue of ethical risks in education has arisen and attracted much attention from academics.Based on the activity-theoretic analysis framework,a systematic review of 150 papers during 2012-2022 on the ethical risks of AI applications in education was conducted,and the results shows that:the research on the ethics of AI in education has gone through three stages:nascent exploration,growth and occurrence,and rapid development;the research topics are mainly safety and privacy,but the research subjects are not clear,and there is a lack of clear direction on the AI that causes ethical problems in education.From a comprehensive perspective,the current research in this field is still in the early stage,and there are more concerns about the ethical dilemmas of AI applications in education,but there is a lack of risk management perspectives,and the existing research results are mostly theoretical and status quo analysis from a single perspective,lacking comprehensive research from multiple theoretical perspectives.The cross-sectional research and multi-perspective theoretical analysis of the ethical risks of AI applications in education are of great significance to the ethics and risk governance of education nowadays.And future research recommendations are proposed:it is necessary to re-examine the interaction between artificial intelligence and education,reshape the ethical norms and evaluation indicators of artificial intelligence in educational applications;The research method is not limited to theoretical research.In the practical teaching of artificial intelligence technology,objective data such as teaching cases and materials are collected,and targeted educational AI ethical governance strategies are proposed scientifically and innovatively;Deeply explore specific issues at different levels and fields of education,and then formulate relevant policies and regulations in a targeted manner,strengthen supervision and management,and enhance the ethical and responsibility awareness of relevant personnel;Not only should we pay attention to the ethical issues involved in the surface application of artificial intelligence education,but we should also pay more attention to the deeper risk occurrence mechanism issues;Strengthen the adjustment of ethical principles and policies for the application of artificial intelligence in education,systematically consider the governance dimensions of artificial intelligence,and strengthen the supervision of education artificial intelligence by the government and relevant management departments by formulating ethical standards.