Abstract:

[Purposes] To retrieve regulations or statements on ethical norms for AI-assisted writing in domestic academic journals, construct a theoretical framework for AI-assisted writing ethics in Chinese journals, and provide references for improving related practices. [Methods] Using CNKI, Wanfang and official wechat public accounts as data sources, 14 academic journal statements were selected, and finally 9 texts were selected for qualitative text analysis using Nvivo20 software. The theoretical model is constructed by combining the Risk Management Guidelines (ISO 31000:2018) and the WHO Guidelines on the Ethics and Governance of Artificial Intelligence in the Field of Health. [Findings] Guided by the six ethical principles of WHO, with authors, editors, and reviewers as the core subjects, and the academic publishing process as the main thread, a theoretical framework for AI-assisted writing ethics in China was established. [Conclusions] The framework of AI-assisted writing ethics provides theoretical support for the application of AI in academic writing, and provides a reference for the practical management of journals. In the future, further research is needed to improve the details of the framework and promote the healthy development of AI-assisted writing.

Key words: AI-assisted writing, Ethical guidelines, Qualitative text analysis, Academic journal

摘要：

【目的】检索国内学术期刊关于人工智能(AI)辅助写作伦理规范的规定或声明,构建我国期刊AI辅助写作伦理规范的理论框架,为完善相关实践提供参考。【方法】以中国知网、万方和官方微信公众号为数据源,筛选出14份学术期刊声明,最终选取9份文本,利用Nvivo20软件进行质性文本分析,并结合《风险管理指南》(ISO 31000:2018)和WHO《卫生健康领域人工智能伦理与治理指南》构建理论模型。【结果】以WHO 6大伦理原则为指引,以作者、编辑、审稿人为核心主体,以学术出版流程为主线,构建了我国AI辅助写作伦理规范的理论框架。【结论】AI辅助写作伦理规范框架为学术写作中AI的应用提供了理论支持,为期刊的实践管理提供了参考。未来须进一步深化研究,完善框架细节,推动AI辅助写作的健康发展。

关键词: 人工智能辅助写作, 伦理规范, 质性文本分析, 学术期刊