透明度(行为)
计算机科学
判断
出版
声誉
工程伦理学
保护
同行评审
质量(理念)
过程(计算)
社会学
知识管理
立场文件
工作流程
个性化
生成语法
信息隐私
公共关系
数据科学
外包
人工智能
德尔菲法
协议(科学)
管理科学
心理学
信息系统
最佳实践
作者
Jan Marco Leimeister,Aleksi Aaltonen,Wendy Currie,Daniel Schlagwein
标识
DOI:10.1177/02683962261458829
摘要
Generative Artificial Intelligence (GenAI) is rapidly entering academic work, including the peer review process. We examine the implications of GenAI for the peer review process – including its use by reviewers, editors, and authors (e.g. for pre-submission self-review or for operationalising review feedback during revision) – and articulates the position of the Journal of Information Technology (JIT). We advance two foundational premises: first, peer review should be seen as a creative process, sometimes even becoming a co-creation with the authors and editors, rather than mechanistic quality control; second, GenAI may augment peer review but should not replace or outsource scholarly judgement and insight. Drawing on recent state-of-the-art analyses of GenAI in peer reviewing, we identify four requirements for ethical use – confidentiality, accountability, bias mitigation, and transparency – and discuss how these principles apply across the reviewing process. We outline where GenAI can provide legitimate support, such as summarisation, language improvement, compliance checking, and workflow management, while emphasising that evaluative judgement must remain with human reviewers and editors. GenAI must not become a shortcut for efficiently producing good papers on average, while steering us away from papers that are more demanding to review but at the same time offering potentially much more impactful contributions. We articulate JIT’s editorial stance for AI-assisted peer review and propose role-specific guidance for authors, reviewers and editors. We also outline a research agenda for studying the impact of GenAI on review quality, timeliness, fairness and trust. Overall, we argue that peer review should evolve through responsible AI augmentation while preserving human-centred governance and the accountability that underpins scholarly evaluation.
科研通智能强力驱动
Strongly Powered by AbleSci AI