Abstract:

[Purposes] This study aims to analyze the legal, ethical, and social issues involved in the works generated by the generative artificial intelligence (AI) model (referred to as AI works) in the academic publishing field, so as to promote the rational application of generative AI in the academic publishing industry. [Methods] This study investigated and sorted out the literature on generative AI models in China and abroad, analyzed the responsibilities and risks of the responsible subjects of AI works, such as authors, AI model developers, and publishers, and initially put forward coping strategies and future innovative publishing modes. [Findings] AI cannot be the author of AI works. Most publishers believe that AI can be used in academic paper writing, but it is necessary to explain the use and utilization method of AI in the appropriate part of the text. AI works may have infringement of copyright and personal rights, ethical dilemmas, ethical conflicts, social injustice, and other social issues. Publishers should insist on manual verification and content review of AI works and work with authors, AI model developers, and regulatory authorities to promote the healthy development and rational application of generative AI. [Conclusions] AI works will inevitably lead to the transformation of the scientific research creation and publishing industry. Publishers must adhere to the manual review of AI works and encourage original output, so as to promote the high-quality publication of scientific research achievements.

Key words: Generative artificial intelligence, Publisher, Author, Legal issue, Ethical risk, Social responsibility, Risk prevention

摘要：

【目的】 分析学术出版领域中生成式人工智能(Artificial Intelligence,AI)模型输出作品(简称“AI作品”)涉及的法律、伦理、社会问题,以促进生成式AI在学术出版行业的合理应用。【方法】 调研与梳理国内外文献资料,分析AI作品相关责任主体如作者、AI模型开发者、出版单位等面临的责任与风险问题,初步提出应对策略与未来创新的出版模式。【结果】 AI不能成为AI作品的作者,多数出版单位认为学术论文创作时可以借助AI,但需在正文适当部分说明是否使用及如何使用AI。AI作品可能存在著作权、人格权等侵权行为,伦理困境、伦理冲突等伦理问题,以及社会不公等社会问题。出版单位应当坚持对AI作品进行人工验证与内容审查,与作者、AI模型开发者、监管部门等共同促进生成式AI的健康发展和规范应用。【结论】 AI作品的出现必将引导科研创作和出版产业发生变革,出版单位必须坚持人工审查,鼓励原创性输出,促进科研成果优质出版。

关键词: 生成式人工智能, 出版单位, 作者, 法律问题, 伦理风险, 社会责任, 风险防范