摘要：

目的 系统探究我国人文社科领域学术期刊对人工智能生成内容（AIGC）的应对策略，揭示其政策制定的阶段性特征、核心共识与实践分化，并为优化我国人文社科期刊AIGC治理生态提供建议。 方法 提出学术期刊AIGC规范策略三级分析框架，结合人工编码与大语言模型技术的混合方法路径，对我国106种人文社科领域中文期刊AIGC使用政策进行量化评估。 结果 我国人文社科期刊的AIGC规范制定呈现从2023年初步响应到2025年常态化跟进的阶段性特征。当前政策体系已形成以作者主体性为核心的共识框架，主要体现在明确作者唯一责任主体、禁止AI署名、将AIGC定位为辅助工具以及确立基础性处理措施四个方面。然而，在具体实践中存在分化，表现为信息披露深度存在差异、审查方式与处理强度不一、规范主体的覆盖范围有限。 结论 当前国内人文社科期刊的AIGC政策正处于从基础性规范向体系化治理的发展阶段。未来应致力于构建覆盖学术出版全流程、多元主体的过程治理体系，推动审查机制向主动化与标准化转变，提升信息披露要求的深度与可追溯性，探索将AIGC治理效能纳入学术期刊评价体系，制定差异化的学科规范，以促进学术出版生态的健康与可持续发展。

关键词: 学术期刊, 人工智能生成内容（AIGC）, 人文社会科学, AIGC使用政策

Abstract:

Purposes The aim of this study is to systematically investigate how academic journals in China’s humanities and social science fields respond to artificial intelligence generated content （AIGC）， to reveal the phased characteristics of policy development， the core consensus and the practice differences， and to provide recommendations for improving the governance ecosystem of AIGC in these journals. Methods A three tier analysis framework of normative strategies for academic journal AIGC was proposed. Using a mixed methods approach that combined manual coding with large language model techniques， we quantitatively assessed the AIGC usage policy texts of 106 Chinese humanities and social science journals. Findings The study found that the development of AIGC norms in Chinese humanities and social science journals follows a phased pattern， from an initial response in early 2023 to routine follow up by 2025. The current policy system has established a consensus framework centered on author agency， reflected mainly in four areas： identifying the sole responsible author， prohibiting AI authorship， positioning AIGC as an auxiliary tool， and establishing basic handling measures. However， in practice there is differentiation， manifested by varying depths of information disclosure， differences in review methods and the intensity of handling measures， and a limited scope of normative subjects. Conclusions Domestic humanities and social science journals’ AIGC policies are progressing from basic norms toward systematic governance. Future efforts should aim to build a process governance system that covers the entire academic publishing workflow and involves multiple stakeholders， promote a proactive and standardized review mechanism， enhance the depth and traceability of information disclosure， explore integrating AIGC governance effectiveness into journal evaluation systems， and develop differentiated discipline norms， thereby fostering a healthy and sustainable academic publishing ecosystem.

Key words: Academic journals, Artificial intelligence generated content （AIGC）, Humanities and social sciences, AIGC usage policies