【摘要】 人工智能（artificial intelligence,AI）的快速发展为我国研究生教育体系带来了前所未有的挑战。理解博士生在学术研究中的AI采用情况，有助于为教育管理者制定有针对性的政策和方案提供依据。为此，本研究基于2024年11449名应届博士毕业生的问卷调查数据，对具有不同特征的博士生在学术写作中采用AI的态度与现状进行了系统分析。研究结果显示：（1）博士生在AI采用意愿与实际使用行为之间存在明显的“转化障碍”，即便是在对AI态度最为积极的群体中，仍有约四成尚未使用AI工具；（2）女性博士生对AI采用普遍呈现出一种“审慎观望”的中间状态；（3）人文、社科博士生在学术成果颇丰的情况下，更倾向于对AI持拒绝态度，其AI使用概率也相应降低，呈现出“学术优势抑制技术使用”的现象。此外，研究进一步印证了理工科背景或重点院校博士生倾向于对AI持积极态度，以及年长博士生在AI使用上相对保守的普遍规律。本研究揭示了博士生在学术写作中采用AI的复杂性，并为研究生教育中的AI素养提升提供了有针对性的建议。更多还原

【Abstract】 The rapid advancement of artificial intelligence(AI) has brought unprecedented challenges to China’s postgraduate education system. Understanding how doctoral students adopt AI in academic writing is therefore essential for informing targeted educational policies and support strategies. Drawing on questionnaire data from 11,449 prospective doctoral graduates in 2024, this study systematically examines group differences in their attitudes toward and use of AI in academic writing. The findings indicate that:(1)there is a “conversion barrier” between doctoral students’ intention to adopt AI and their actual use. Even among those with the most positive attitudes, approximately 40% have not yet used AI to assist academic writing;(2) female doctoral students tend to adopt a cautious, wait-and-see orientation toward AI adoption; and(3) doctoral students in the humanities and social sciences, despite relatively strong academic output, are more inclined to reject AI and are less likely to use it, revealing a pattern whereby academic advantage suppresses technology use. In addition, the study confirms empirical observations from existing research: doctoral students in STEM disciplines or from top-tier universities are generally more positive toward AI, whereas older doctoral students tend to be more conservative in AI adoption. Overall, this study highlights the complexity of AI adoption in doctoral academic writing and offers targeted, evidence-informed implications for enhancing AI literacy in postgraduate education.更多还原