中国电化教育

爆火出圈的ChatGPT为AIGC的发展带来了重大突破，也使教育领域从普适走向个性、从单一走向多样、从零散走向整合，人工智能赋能教育进入新阶段，随之也面临AIGC带来的一系列伦理风险，引起教育工作者的高度关注，需要从风险性监管视角来处理AIGC教育应用的伦理风险。该文首先从“人-技术”交互角度归纳AIGC教育应用四类伦理风险：算法相关伦理风险、数据相关伦理风险、技术相关伦理风险和应用相关伦理风险。其次以风险性监管流程为基础制定AIGC教育应用伦理风险的治理框架，依据治理框架对四类伦理风险按照“归纳-运行-分级”的顺序实行监管：在总结伦理风险已有的治理依据后构建“5W1H”监管模型，对伦理风险进行编码后实现高、中、低三类分级。以此系统化的监管流程为应对AIGC教育应用伦理风险提供借鉴，对AIGC时代赋能教育的伦理使用指南提供指向性的引领，推动教育领域这场持久变革的健康化发展。

ChatGPT, which bursts out of the circle, has brought a major breakthrough in the development of AIGC, and also made the field of education move from universal to individuality, from single to diversity, from fragmentation to integration, and AIenabled education enter a new stage, and then also face a series of ethical risks brought by AIGC, which arouses great concern among educators and requires a risky regulatory perspective to deal with the ethical application of AIGC education. Risks. Firstly, the four types of ethical risks of AIGC educational applications are summarized from the perspective of “human-technology” interaction:algorithm-related ethical risks, data-related ethical risks, technology-related ethical risks and application-related ethical risks.Secondly, we formulate a governance framework for ethical risks of AIGC educational applications based on the risk-based regulatory process, and implement the regulation of the four types of ethical risks in the order of “generalization-operation-grading” based on the governance framework: after summarizing the existing governance basis for ethical risks, we construct the “5W1H” regulatory model is constructed after summarizing the existing ethical risk management basis, and the ethical risks are coded and graded into three categories: high, medium and low. This systematic regulatory process provides a reference for addressing the ethical risks of AIGC educational applications, and provides guidance for the ethical use of empowered education in the AIGC era to promote the healthy development of this lasting change in education.