国际出版机构使用人工智能工具的出版伦理政策分析

Analysis on Publishing Ethics Policies of International Publishers for the Use of Artificial Intelligence Tools

从科学网(Web of Science)的期刊引证报告(Journal Citation Reports)数据库排名前70的出版机构中筛选出23家,分析其人工智能(AI)出版伦理政策。在作者使用AI方面,出版机构主要从署名、AI使用边界、披露形式进行了规范。出版机构对AI身份问题已达成共识,即AI不能成为署名作者;出版机构允许在研究和写作中使用AI,但使用者需要对内容负责且充分披露,而目前只有少数(26.1%)机构明确规定了作者的AI使用边界,且各机构之间对于如何披露的规定存在差异。关于审稿者使用AI的政策,鉴于保密性要求,机构一般不允许审稿者将稿件上传至无法保证安全性的AI工具,但关于如何披露和违反AI使用政策的惩罚措施还有待完善。针对上述问题,建议出版机构从伦理、技术和法律三个层面,以及协同治理的角度,与学术共同体和管理部门合作,共同制定学术出版中AI使用的伦理规范,以营造健康的学术生态环境。

The article selectes 23 publishing institutions from the top 70 in the Journal Citation Reports database of Web of Science and analyzes their artificial intelligence(AI)publishing ethics policies.In terms of the author’s use of AI,publishing institutions have mainly standardized the authorship,AI usage boundaries and disclosure forms.Publishing institutions have reached a consensus on the issue of AI identity.That is,AI cannot become a coauthor.Publishing institutions allow the use of AI in research and writing,but require authors’responsibility and full disclosure.Currently only a few institutions(26.1%)have clearly defined the boundaries of authors’AI usage,and there are differences among institutions on how to disclose them.Regarding the policy on the use of AI by reviewers,due to confidentiality requirements,institutions generally do not allow reviewers to upload manuscripts to AI tools that cannot guarantee security.However,there is still room for improvement on how to disclose and punish violations of AI usage policies.In response to the above issues,it is suggested that publishing institutions collaborate with academic communities and management departments from the perspectives of ethics,technology and law,as well as collaborative governance,to jointly develop ethical standards for AI applications in academic publishing,in order to create a healthy academic ecosystem.