Rethinking Higher Education/Chapter 10/en-zh

Language: EN · ZH · EN-ZH · ← Book

Chapter 10: University of the Future

Martin Woesler

English (Source)	中文 (Target)
== University of the Future: AI-Enhanced Higher Education Between European Humanism and Chinese Innovation ==	== 未来大学：欧洲人文主义与中国创新之间的人工智能增强高等教育 ==
Martin Woesler	Martin Woesler
Hunan Normal University	湖南师范大学
Abstract	摘要
Higher education stands at a crossroads. The convergence of artificial intelligence, post-pandemic hybrid learning, and smart campus technologies is transforming universities from knowledge-transmission institutions into adaptive learning ecosystems. This article examines how European and Chinese universities are responding to this transformation, drawing on institutional data, policy analysis, and recent empirical research. We document China‘s state-led approach — exemplified by the UNESCO-award-winning Smart Education Platform serving 293 million learners — and contrast it with the EU’s decentralized model operating through 65 European Universities alliances encompassing over 570 institutions across 35 countries. Through a systematic comparison of AI adoption policies, hybrid learning models, smart campus infrastructure, and generative AI integration, we argue that neither the Chinese emphasis on speed and scale nor the European emphasis on democratic governance and faculty autonomy is sufficient alone. A synthesis — combining China’s capacity for rapid deployment with Europe’s commitment to humanistic values and institutional self-governance — offers the most promising model for the university of the future.	高等教育正处于十字路口。人工智能、后疫情时代混合学习与智慧校园技术的融合，正在将大学从知识传递机构转变为自适应学习生态系统。本文考察了欧洲和中国的大学如何应对这一转型，借助机构数据、政策分析和近期实证研究。我们记录了中国的国家主导模式——以荣获联合国教科文组织奖的智慧教育平台（服务2.93亿学习者）为代表——并将其与欧盟通过65个欧洲大学联盟（涵盖35个国家570余所院校）运作的分散化模式进行对比。通过对人工智能采纳政策、混合学习模式、智慧校园基础设施和生成式人工智能整合的系统比较，我们认为，无论是中国对速度和规模的强调，还是欧洲对民主治理和教师自治的重视，单独而言均不充分。将中国的快速部署能力与欧洲对人文主义价值观和机构自治的承诺相结合的综合方案，为未来大学提供了最具前景的模式。
Keywords: university transformation, AI in higher education, smart campus, hybrid learning, EU-China comparison, generative AI policy, digital education	关键词：大学转型、高等教育中的人工智能、智慧校园、混合学习、欧中比较、生成式人工智能政策、数字教育
1. Introduction	1. 引言
On 13 March 2024, the Jean Monnet Centre of Excellence at Hunan Normal University inaugurated its lecture series „Digitalization in China and Europe“ with a presentation entitled „University of the Future.“ The lecture, attended by more than 200 participants, explored how artificial intelligence, virtual reality, and hybrid pedagogies are reshaping the institution that has served as the primary vehicle for advanced knowledge transmission since the founding of the University of Bologna in 1088.	2024年3月13日，湖南师范大学让·莫内卓越中心以"未来大学"为主题的演讲，开启了"中欧数字化"系列讲座。该讲座吸引了200余名参与者，探讨了人工智能、虚拟现实和混合教学法如何重塑这一自1088年博洛尼亚大学创建以来一直作为高等知识传递主要载体的机构。
The university faces a paradox. As an institution, it is remarkably stable — its basic organizational form (departments, faculties, lectures, examinations, degrees) has changed less in a millennium than almost any other social institution. Yet the environment in which it operates has changed beyond recognition within a single decade. Students entering university today will graduate into a labour market in which an estimated 44 percent of workers will need to change their skill profile within five years (WEF 2023). They will use AI tools daily — 50 percent already use them at least weekly inside and outside the classroom (EDUCAUSE 2025). They expect hybrid learning options — 86 percent of residential undergraduate students prefer some combination of in-person and online courses (Rize Education 2025). And they will work in an economy where the institutions that train them — corporations, online platforms, bootcamps — increasingly compete with universities for the credential market.	大学面临一个悖论。作为一种机构，它异常稳定——其基本组织形式（系、学院、授课、考试、学位）在千年间的变化比几乎任何其他社会机构都小。然而，其运营环境在十年之内已经面目全非。今天入学的大学生将步入一个劳动力市场——据估计，44%的劳动者需要在五年内改变其技能结构（WEF 2023）。他们将每天使用人工智能工具——50%的学生已在课堂内外至少每周使用一次（EDUCAUSE 2025）。他们期待混合学习选择——86%的住校本科生倾向于面授与在线课程的某种组合（Rize Education 2025）。而他们将在一个培训机构——企业、在线平台、训练营——日益与大学争夺资质市场的经济体中工作。
This article examines how European and Chinese universities are navigating this transformation. It builds on the companion chapters in this volume that address specific dimensions of educational digitalization — AI in language learning, alternative learning forms, data protection, AI ethics, and sustainability — by addressing the overarching institutional question: what kind of university do we need for the AI age, and how are the world’s two largest educational systems working toward it?	本文考察了欧洲和中国大学如何驾驭这一转型。它建立在本论文集配套章节——涉及语言学习中的人工智能、替代学习形式、数据保护、人工智能伦理和可持续性——的基础之上，回答总体性的机构问题：人工智能时代我们需要什么样的大学，世界两大教育体系如何朝着这一目标努力？
2. The Smart Campus: Infrastructure for the Future	2. 智慧校园：未来的基础设施
2.1 China‘s Smart Education Platform	2.1 中国的智慧教育平台
China‘s approach to university transformation is exemplified by the National Smart Education Platform (国家智慧教育平台), launched in March 2022 and recognized with the UNESCO King Hamad Bin Isa Al-Khalifa Prize for the Use of ICT in Education in 2023. By the end of 2023, the platform linked 519,000 educational institutions, 18.8 million teachers, and 293 million learners, with over 100 million registered users from more than 200 countries and 36.7 billion visits (UNESCO 2023).	中国的大学转型方法以国家智慧教育平台（国家智慧教育平台）为代表，于2022年3月推出，2023年荣获联合国教科文组织哈马德·本·伊萨·阿勒哈利法国王信息通信技术教育奖。截至2023年底，该平台连接了519,000个教育机构、1,880万名教师和2.93亿名学习者，拥有来自200多个国家的超过1亿注册用户，访问量达367亿次（UNESCO 2023）。
The platform integrates four sub-platforms — for basic education, vocational education, higher education, and teacher training — into a unified digital infrastructure. For higher education specifically, it provides access to more than 27,000 online courses, virtual simulation experiments, and inter-institutional collaborative learning spaces. The scale is unprecedented: no other country operates a single educational platform serving nearly 300 million users.	该平台将四个子平台——基础教育、职业教育、高等教育和教师培训——整合到统一的数字基础设施中。具体到高等教育，它提供了超过27,000门在线课程、虚拟仿真实验和跨校协作学习空间的访问。其规模史无前例：没有其他国家运营着服务近3亿用户的单一教育平台。
The Smart Education Platform reflects a distinctly Chinese approach to educational technology — centralized, state-funded, rapidly deployed, and integrated with broader national strategies. The 14th Five-Year Plan’s education digitalization strategy positions digital transformation as essential to China‘s goal of becoming a global education leader by 2035. Provincial governments have supplemented the national platform with regional initiatives; Hunan Province, for example, has invested in smart classroom infrastructure across its public universities, including Hunan Normal University.	智慧教育平台体现了一种典型的中国教育技术路径——集中化、国家资助、快速部署，并与更广泛的国家战略相整合。"十四五"规划的教育数字化战略将数字化转型定位为中国到2035年成为全球教育强国目标的关键要素。各省级政府以地方倡议补充了国家平台；例如，湖南省已在其公立大学（包括湖南师范大学）投资建设智慧教室基础设施。
2.2 European Smart Campus Initiatives	2.2 欧洲智慧校园倡议
The European approach to smart campus development is characteristically decentralized. Rather than a single national or continental platform, European universities pursue smart campus initiatives individually or through collaborative networks. The European Universities Initiative, funded through Erasmus+, has established 65 alliances encompassing over 570 higher education institutions across 35 countries (European Commission 2025). These alliances facilitate shared digital infrastructure, joint online programmes, and collaborative research, but each institution retains autonomy over its technological choices and pedagogical approaches.	欧洲的智慧校园发展方法具有典型的分散化特征。欧洲大学并非依托单一的国家或大陆平台，而是以个体或通过协作网络的方式推进智慧校园建设。欧洲大学倡议通过Erasmus+资助，建立了65个联盟，涵盖35个国家的570余所高等教育机构（European Commission 2025）。这些联盟促进了共享数字基础设施、联合在线项目和协作研究，但每所院校在技术选择和教学方法上保留自主权。
Individual European universities have developed notable smart campus projects. The University of Edinburgh’s Smart Data Campus integrates IoT sensors, learning analytics, and predictive systems. TU Delft’s Smart Campus initiative uses digital twin technology to optimize building management and learning environments. The European Commission’s Digital Education Action Plan (2021-2027) provides policy guidance and funding incentives but, unlike China‘s approach, does not mandate specific technological solutions.	个别欧洲大学开发了值得关注的智慧校园项目。爱丁堡大学的智慧数据校园整合了物联网传感器、学习分析和预测系统。代尔夫特理工大学的智慧校园倡议使用数字孪生技术来优化建筑管理和学习环境。欧盟委员会的《数字教育行动计划》（2021—2027年）提供了政策指导和资金激励，但与中国的模式不同，并不强制要求采用特定的技术方案。
The EU approach offers advantages in institutional diversity and faculty autonomy — universities can adapt technologies to their specific contexts and pedagogical traditions. However, it also produces fragmentation, duplication of effort, and slower adoption compared to China‘s centralized model.	欧盟模式在机构多样性和教师自治方面具有优势——大学可以根据自身特定情境和教学传统调适技术。然而，与中国的集中化模式相比，它也导致了碎片化、重复建设以及更缓慢的采纳速度。
2.3 Smart Classroom Design and Learning Outcomes	2.3 智慧课堂设计与学习成效
Research on smart classroom effectiveness provides empirical grounding for institutional investment decisions. A study of 421 Chinese undergraduates from Project 985, Project 211, and local universities found that „psychological enjoyment triggered by immersive smart-classroom infrastructure is an important source of perceived academic improvement“ and that „teacher-directed scaffolding with AI amplifies this effect“ (Zhang, C. 2026). The study suggests that smart classroom technology enhances learning not primarily through information delivery but through the creation of engaging, immersive environments that increase student motivation and attention.	关于智慧课堂效果的研究为机构投资决策提供了实证依据。一项对来自985工程、211工程和地方院校的421名中国本科生的研究发现，"沉浸式智慧课堂基础设施所激发的心理愉悦感是感知学业提升的重要来源"，且"教师引导的人工智能支架效应放大了这一效果"（Zhang, C. 2026）。该研究表明，智慧课堂技术提升学习效果的途径主要不是通过信息传递，而是通过营造引人入胜的沉浸式环境来提高学生的学习动机和注意力。
A systematic review of smart campus technologies identified the integration of IoT, AI, cloud computing, and Big Data analytics as the key elements of intelligent campus infrastructure, with personalized dashboards, AI chatbots, and predictive analytics emerging as the most promising applications (Elbertsen, Kok, & Salimi 2025). However, the review also noted that most smart campus implementations remain at the level of facilities management rather than pedagogical transformation — optimizing building energy use rather than fundamentally changing how teaching and learning occur.	一项关于智慧校园技术的系统综述确定了物联网、人工智能、云计算和大数据分析的整合为智能校园基础设施的关键要素，其中个性化仪表板、人工智能聊天机器人和预测分析被认为是最有前景的应用（Elbertsen, Kok & Salimi 2025）。然而，该综述也指出，大多数智慧校园的实施仍停留在设施管理层面而非教学变革层面——即优化建筑能耗而非从根本上改变教学方式。
3. AI-Personalized Learning	3. 人工智能个性化学习
3.1 The Promise and the Reality	3.1 愿景与现实
The aspiration for AI-personalized learning — instruction tailored to each student’s pace, style, prior knowledge, and learning objectives — represents one of the most compelling visions for the university of the future. China has embraced this vision with particular enthusiasm. The Smart Education Platform incorporates adaptive learning algorithms that recommend courses, adjust difficulty levels, and provide individualized feedback based on student performance data.	人工智能个性化学习的愿景——根据每位学生的进度、风格、先验知识和学习目标量身定制教学——代表了未来大学最引人注目的构想之一。中国以尤为饱满的热情拥抱了这一愿景。智慧教育平台融入了自适应学习算法，可根据学生学业数据推荐课程、调整难度等级并提供个性化反馈。
However, empirical research reveals a more nuanced picture. A thematic analysis of 48 Chinese undergraduates’ perceptions of AI-personalized learning found that students valued efficiency gains — faster access to relevant materials, more targeted practice exercises, immediate feedback — but expressed concerns about over-reliance on AI recommendations and the potential loss of critical thinking and autonomous learning capacity (Wang et al. 2024). Students described a tension between the convenience of AI-guided learning paths and their desire to explore, struggle, and discover independently.	然而，实证研究揭示了更为细致的图景。一项对48名中国本科生关于人工智能个性化学习感知的主题分析发现，学生重视效率提升——更快获取相关材料、更有针对性的练习题、即时反馈——但对过度依赖人工智能推荐以及批判性思维和自主学习能力的潜在丧失表达了担忧（Wang等 2024）。学生描述了人工智能引导学习路径的便利性与自主探索、挣扎和发现的渴望之间的张力。
This tension reflects a deeper philosophical question about education. If the purpose of university education is merely to transmit predetermined knowledge and skills as efficiently as possible, then AI personalization represents an unequivocal improvement. If, however, education is also about developing intellectual autonomy, tolerance for ambiguity, the capacity for critical thinking, and the ability to formulate questions rather than merely answer them, then excessive personalization may be counterproductive — it may optimize the measurable while undermining the essential.	这一张力折射出关于教育的更深层哲学问题。如果大学教育的目的仅仅是尽可能高效地传递预定知识和技能，那么人工智能个性化代表了无可争议的改进。但如果教育还在于培养智识自主、对模糊性的包容、批判性思维的能力以及提出问题而非仅仅回答问题的能力，那么过度个性化可能适得其反——它可能优化了可量化的方面，同时削弱了本质性的方面。
3.2 Personalization in Practice: Three Models	3.2 个性化的实践：三种模式
Three distinct models of AI-personalized learning have emerged in current practice:	当前实践中已经出现了三种不同的人工智能个性化学习模式：
The recommendation model uses AI to suggest courses, readings, and learning activities based on student performance data and preferences. This is the approach of most MOOC platforms and of China‘s Smart Education Platform. It treats education as analogous to content consumption — students are „recommended“ educational experiences in the same way that Netflix recommends films. The model is efficient for self-directed learning but raises concerns about filter bubbles (students being channelled toward comfortable material rather than challenging content) and about reducing education to consumption.	推荐模式利用人工智能根据学生的学业数据和偏好推荐课程、阅读材料和学习活动。这是大多数MOOC平台和中国智慧教育平台采用的方法。它将教育类比为内容消费——学生被"推荐"教育体验，正如Netflix推荐电影一样。该模式对自主学习而言效率较高，但引发了对"信息茧房"（学生被引导至舒适材料而非挑战性内容）以及将教育简化为消费的担忧。
The adaptive model adjusts the difficulty, pace, and sequence of learning materials in real time based on student responses. Adaptive learning platforms such as Knewton, DreamBox, and China‘s Squirrel AI have demonstrated measurable improvements in standardized test performance. However, critics argue that adaptive learning optimizes for measurable outcomes while potentially neglecting the unmeasurable — curiosity, creativity, the willingness to engage with difficulty rather than routing around it.	自适应模式根据学生的实时反馈调整学习材料的难度、进度和序列。Knewton、DreamBox等自适应学习平台以及中国的松鼠Ai已在标准化考试成绩方面取得了可衡量的改进。然而，批评者指出，自适应学习针对可量化成果进行优化，却可能忽视了不可量化的方面——好奇心、创造力、直面困难而非绕道而行的意愿。
The tutoring model uses AI chatbots or virtual tutors to provide individualized instruction, answering questions, explaining concepts, and guiding problem-solving. Studies of AI tutoring in mathematics and science have shown promising results, particularly for students who might not otherwise have access to individual tutoring. The AI-assisted language learning study in this volume found that AI tutoring provides psychological benefits — particularly the „no fear of making mistakes“ effect — that supplement rather than replace human instruction (Woesler, this volume).	辅导模式使用人工智能聊天机器人或虚拟导师提供个性化教学，回答问题、解释概念并指导问题解决。关于人工智能辅导在数学和科学领域的研究显示了令人鼓舞的成果，尤其对于原本无法获得个别辅导的学生而言。本论文集中关于人工智能辅助语言学习的研究发现，人工智能辅导提供了心理层面的益处——特别是"不怕犯错"的效应——这补充而非取代了人类教学（Woesler, 本卷）。
Each model has legitimate applications, but none constitutes „personalized education“ in the deeper sense envisioned by educational philosophers. Genuine personalization would involve not merely adapting content delivery to individual learning patterns but supporting each student’s development as a unique intellectual and moral agent — a goal that remains beyond current AI capabilities.	每种模式都有其合理的应用场景，但没有一种构成教育哲学家所设想的更深层意义上的"个性化教育"。真正的个性化不仅仅是将内容交付适配于个体学习模式，而是支持每位学生作为独特智识和道德主体的发展——这一目标仍然超出了当前人工智能的能力范围。
3.3 European Caution and Chinese Enthusiasm	3.3 欧洲的审慎与中国的热忱
European institutions have generally approached AI-personalized learning with greater caution than their Chinese counterparts. This reflects several factors: stronger data protection frameworks (the GDPR imposes strict requirements on the processing of student data for personalization purposes; see the companion chapter on data protection, Woesler, this volume), a tradition of faculty autonomy in pedagogical decisions, and a humanistic educational philosophy that emphasizes Bildung — the formation of the whole person — over skill acquisition.	欧洲机构对人工智能个性化学习的态度总体上比中国同行更为审慎。这反映了几方面因素：更强的数据保护框架（GDPR对出于个性化目的处理学生数据施加了严格要求；参见本卷配套的数据保护章节，Woesler）、教师在教学决策中享有自主权的传统，以及强调教养（Bildung）——即全人培养——优先于技能习得的人文主义教育哲学。
Chinese institutions, operating within a more permissive data environment and encouraged by state policy, have moved more aggressively toward AI personalization. The Ministry of Education’s emphasis on „smart education“ (智慧教育) and the integration of AI into the National Smart Education Platform create institutional incentives for adoption. However, the empirical evidence on effectiveness remains mixed, and Chinese researchers are increasingly calling for critical evaluation alongside enthusiasm.	中国的机构在更为宽松的数据环境中运作，并受到国家政策的鼓励，因而更积极地推进人工智能个性化。教育部对"智慧教育"（智慧教育）的强调以及人工智能在国家智慧教育平台中的整合，为采纳创造了制度激励。然而，关于有效性的实证证据仍然参差不齐，中国研究者也日益呼吁在热忱之外进行批判性评估。
4. Hybrid Learning as Post-Pandemic Standard	4. 混合学习作为后疫情时代的标准
4.1 The Durability of Hybrid Models	4.1 混合模式的持久性
The COVID-19 pandemic forced universities worldwide to adopt online learning at scale. The question in 2024-2026 is not whether to continue some form of online delivery but how to optimize the blend. Data suggest that hybrid learning is not a temporary accommodation but a permanent feature of higher education. Over 68 percent of universities have increased their online offerings since 2020, with nearly half planning to make online programmes a central strategic pillar (EDUCAUSE 2025). A survey of residential undergraduate students found that 86 percent prefer a hybrid approach with one to four online courses per semester (Rize Education 2025).	新冠疫情迫使全球大学大规模采用在线学习。2024至2026年间的问题不是是否继续某种形式的在线教学，而是如何优化混合方式。数据表明，混合学习不是临时安排，而是高等教育的永久特征。超过68%的大学自2020年以来增加了在线教学供给，近半数计划将在线项目作为核心战略支柱（EDUCAUSE 2025）。一项对住校本科生的调查发现，86%的学生偏好每学期包含一至四门在线课程的混合模式（Rize Education 2025）。
Research using Structural Equation Modeling and Deep Neural Network analysis confirmed that hybrid learning represents a sustainable post-pandemic model, though its effectiveness depends heavily on institutional support infrastructure — including reliable technology, trained faculty, appropriate assessment design, and student support services (Yaqin et al. 2025). The study identified institutional support as a stronger predictor of hybrid learning success than either technology quality or student digital competence.	使用结构方程模型和深度神经网络分析的研究证实，混合学习代表了一种可持续的后疫情模式，但其有效性在很大程度上取决于机构支持基础设施——包括可靠的技术、经过培训的教师、适当的评估设计和学生支持服务（Yaqin等 2025）。该研究发现，机构支持比技术质量或学生数字素养更能预测混合学习的成功。
4.2 Chinese Blended Learning Adoption	4.2 中国的混合学习采纳
In China, a study of factors influencing behavioural intention to adopt blended learning among university students in the post-pandemic era found broad acceptance alongside specific concerns (Yu et al. 2023). Students valued the flexibility of blended learning — the ability to review recorded lectures, access materials asynchronously, and study at their own pace — but expressed concern about reduced social interaction and the difficulty of maintaining self-discipline in online environments.	在中国，一项关于后疫情时代大学生混合学习采纳行为意愿影响因素的研究发现了广泛的接受与具体的担忧并存（Yu等 2023）。学生重视混合学习的灵活性——回看录播课程、异步获取材料、按自身节奏学习——但对社交互动减少以及在线环境中维持自律的困难表示担忧。
These concerns resonate with longstanding critiques of online education. The social dimension of university learning — the informal conversations in corridors, the debates in seminars, the collaborative projects that build professional networks — cannot be fully replicated in digital environments. Chinese universities have responded by developing hybrid formats that combine online content delivery with intensive face-to-face workshops, collaborative projects, and mentoring sessions. Hunan Normal University’s own experience with the Jean Monnet lecture series illustrates this approach: lectures were delivered to audiences of 150-200 participants with simultaneous online streaming, followed by in-person discussion sessions with smaller groups.	这些担忧与关于在线教育的长期批评不谋而合。大学学习的社交维度——走廊里的非正式对话、研讨课上的辩论、建立专业网络的合作项目——无法在数字环境中得到完全复制。中国大学的应对方式是开发混合格式，将在线内容交付与密集的面授工作坊、合作项目和辅导环节相结合。湖南师范大学自身的让·莫内系列讲座经验便是这一路径的体现：讲座面向150至200名现场参与者进行，同时进行在线直播，随后安排小组面授讨论环节。
A meta-analysis covering 37 blended learning studies from 2000 to 2024 found a positive upper-medium effect on learning outcomes (SMD = 0.698), with the optimal online proportion at approximately 50 percent. This finding is significant for curriculum design: it suggests that the most effective hybrid model is not primarily online with occasional face-to-face sessions, nor primarily face-to-face with supplementary online materials, but a roughly equal blend of both. The pedagogical implication is that hybrid learning should be designed as an integrated experience — not simply „the same course delivered through two channels“ — with specific learning activities assigned to the mode best suited to their educational purpose: information transmission online, social learning and deep discussion in person.	一项涵盖2000至2024年37项混合学习研究的荟萃分析发现了对学习成效的上中等正向效应（SMD = 0.698），最佳在线比例约为50%。这一发现对课程设计具有重要意义：它表明最有效的混合模式既不是以在线为主偶尔面授，也不是以面授为主辅以在线材料，而是两者大致对等的混合。其教学启示在于，混合学习应被设计为一种整合化体验——而非简单的"同一门课程通过两种渠道交付"——将特定学习活动分配至最适合其教育目的的模式：信息传递在线进行，社交学习和深度讨论则在面授中进行。
5. Generative AI in the Curriculum	5. 课程中的生成式人工智能
5.1 The Adoption Wave	5.1 采纳浪潮
Generative AI has moved from novelty to ubiquity in higher education with remarkable speed. The EDUCAUSE 2025 survey found that 57 percent of higher education institutions now prioritize AI integration, up from 49 percent in 2024. A global analysis of institutional adoption policies found a spectrum of responses, from outright bans (increasingly rare) to mandatory integration (still uncommon) to the emerging mainstream of „regulated integration“ — permitting AI use under specified conditions with appropriate attribution (Computers and Education: Artificial Intelligence 2025).	生成式人工智能以惊人的速度从新事物转变为高等教育中的普遍现象。EDUCAUSE 2025调查发现，57%的高等教育机构现在优先考虑人工智能整合，高于2024年的49%。一项关于机构采纳政策的全球分析发现了从完全禁止（日益罕见）到强制整合（仍不常见）到"规范化整合"这一新兴主流——在规定条件下允许使用人工智能并要求适当标注（Computers and Education: Artificial Intelligence 2025）。
In the United States, 63 percent of high-research-activity universities encourage generative AI use, while 27 percent discourage or limit it. In Europe, policies vary dramatically by institution and country, reflecting the decentralized governance model. In China, the picture is complicated by the restriction of access to global AI tools (ChatGPT, Claude, Gemini, Copilot) and the promotion of domestic alternatives (Ernie Bot, Tongyi Qianwen, DeepSeek).	在美国，63%的高研究活跃度大学鼓励使用生成式人工智能，而27%予以限制或禁止。在欧洲，政策因机构和国家不同而差异显著，反映了分散化的治理模式。在中国，情况因限制访问全球人工智能工具（ChatGPT、Claude、Gemini、Copilot）以及推广国内替代品（文心一言、通义千问、DeepSeek）而更为复杂。
5.2 Chinese Students Navigating Global and Domestic AI	5.2 中国学生驾驭全球与国内人工智能
A study of how Chinese university students navigate global and domestic generative AI found that students develop complex strategies to access restricted tools alongside officially permitted domestic alternatives (Xie et al. 2025). Students reported using VPNs, foreign accounts, and peer networks to access ChatGPT and other restricted platforms, while also using domestic tools for assignments that would be reviewed by instructors. This dual-use pattern reflects a tension in China‘s AI education strategy between promoting AI adoption and controlling the AI ecosystem.	一项关于中国大学生如何驾驭全球和国内生成式人工智能的研究发现，学生发展了复杂的策略来获取受限工具，同时使用官方允许的国内替代品（Xie等 2025）。学生报告使用VPN、境外账户和同伴网络来访问ChatGPT和其他受限平台，同时使用国内工具完成将被教师审阅的作业。这种双轨使用模式反映了中国人工智能教育战略中推动人工智能采纳与管控人工智能生态之间的张力。
5.3 Academic Integrity Challenges	5.3 学术诚信挑战
The integration of generative AI raises pressing questions about academic integrity. A study found that 58 percent of students admitted to using AI to complete assignments dishonestly, while 65 percent acknowledged that plagiarism is a concern with unmodified AI content (ICAI 2024; Frontiers in Artificial Intelligence 2024). Chinese universities have responded with quantitative measures — Tianjin University of Science and Technology (天津科技大学), for example, mandated in 2024 that AI-generated content in undergraduate theses must not exceed 40 percent. The Ministry of Education has prohibited students from submitting AI-generated content as their own academic work.	生成式人工智能的整合提出了关于学术诚信的紧迫问题。一项研究发现，58%的学生承认曾不诚实地使用人工智能完成作业，65%的学生认为未经修改的人工智能内容存在抄袭风险（ICAI 2024; Frontiers in Artificial Intelligence 2024）。中国大学以定量措施予以回应——例如，天津科技大学（天津科技大学）于2024年规定本科论文中人工智能生成的内容不得超过40%。教育部已禁止学生将人工智能生成的内容作为自己的学术成果提交。
These responses illustrate the difficulty of regulating AI use in education. A 40 percent threshold is measurable in principle but difficult to enforce in practice — how does one determine the exact proportion of AI-generated content in a text that has been substantially rewritten by its human author? The deeper challenge is not detecting AI use but developing assessment methods that value the distinctly human contributions — original thinking, critical analysis, creative synthesis — that AI cannot replicate. This is where the philosophical traditions of European Bildung and Chinese 修身 (xiushen, self-cultivation) converge: both emphasize the development of the whole person, not merely the production of correct answers.	这些回应说明了在教育中规范人工智能使用的困难。40%的门槛原则上可衡量，但在实践中难以执行——如何确定一篇经人类作者大幅改写的文本中人工智能生成内容的确切比例？更深层的挑战不是检测人工智能的使用，而是开发重视独特人类贡献——原创思维、批判性分析、创造性综合——的评估方法，这些是人工智能无法复制的。这正是欧洲教养（Bildung）和中国修身（xiushen）哲学传统的交汇之处：两者都强调全人的发展，而非仅仅产出正确答案。
5.4 Institutional Policy Landscapes	5.4 机构政策格局
The global picture of GenAI policy in higher education is characterized by rapid evolution and significant institutional variation. In China, Fudan University became one of the first major Chinese universities to issue comprehensive AI guidelines in January 2026, addressing both student and teacher use. The guidelines permit AI for research assistance and learning support but prohibit its use in formal assessments without explicit instructor authorization. Other Chinese universities have followed with varying degrees of permissiveness, though the overarching direction — shaped by the Ministry of Education — is toward regulated integration rather than prohibition.	高等教育中生成式人工智能政策的全球图景以快速演变和显著的机构差异为特征。在中国，复旦大学成为2026年1月首批发布全面人工智能指南的中国重点大学之一，涵盖学生和教师的使用。该指南允许将人工智能用于研究辅助和学习支持，但禁止在未经教师明确授权的情况下用于正式考核。其他中国大学相继跟进，宽严程度各异，但总体方向——在教育部的引导下——是走向规范化整合而非禁止。
In Europe, the OECD Digital Education Outlook (2023) found that none of the 18 countries or jurisdictions surveyed had issued specific regulations on generative AI in education, though nine had published non-binding guidance. By 2025, this picture had evolved significantly, with several countries (including the Netherlands, Finland, and Ireland) issuing national guidelines and individual universities developing institution-specific policies. However, the pace of policy development continues to lag behind the pace of AI adoption by students and faculty.	在欧洲，OECD《数字教育展望》（2023）发现，所调查的18个国家或地区中没有一个出台了关于教育中生成式人工智能的专门法规，但有9个发布了非约束性指导。到2025年，这一局面已有显著变化，数个国家（包括荷兰、芬兰和爱尔兰）发布了国家指南，各大学也制定了机构层面的政策。然而，政策制定的步伐仍然落后于学生和教师采纳人工智能的速度。
The policy landscape in higher education increasingly shows a readiness gap. The EDUCAUSE 2024 AI Landscape Study found that 80 percent of faculty and staff use AI tools, yet fewer than one in four are aware of their institution’s formal AI policy. This gap between practice and governance represents one of the most urgent challenges for the university of the future — institutional policies must not only exist but be communicated, understood, and perceived as relevant by the academic community.	高等教育中的政策格局日益呈现出"准备度差距"。EDUCAUSE 2024人工智能态势研究发现，80%的教职员工使用人工智能工具，但不到四分之一的人了解其所在机构的正式人工智能政策。实践与治理之间的这一差距是未来大学面临的最紧迫挑战之一——机构政策不仅要存在，还要被传达、被理解，并被学术共同体视为切实相关。
6. The Role of Faculty in University Transformation	6. 教职人员在大学转型中的角色
6.1 Faculty as Agents of Change	6.1 教职人员作为变革推动者
The transformation of the university cannot be imposed from above — it must be led, or at least co-led, by faculty. A critical reflection on current AI trends in higher education argued that transformation must be „faculty-led rather than technology-driven“ (Ruano-Borbalan 2025). Faculty members are the carriers of disciplinary knowledge, the designers of curricula, and the primary interface between students and the institution. Their engagement — or resistance — determines whether technological innovation translates into genuine pedagogical improvement.	大学的转型不能自上而下强加——它必须由教职人员领导或至少共同领导。一篇关于高等教育中人工智能当前趋势的批判性反思认为，转型必须是"教职人员主导而非技术驱动"的（Ruano-Borbalan 2025）。教职人员是学科知识的承载者、课程的设计者，以及学生与机构之间的主要桥梁。他们的参与——或抵触——决定了技术创新能否转化为真正的教学改进。
Yet faculty face significant challenges. Many academics trained before the AI era lack personal experience with the tools their students use daily. The incentive structures of academic careers — which reward research publications over teaching innovation in most systems — provide limited motivation for investing time in pedagogical transformation. And the speed of technological change creates uncertainty about which innovations are worth adopting and which are ephemeral trends.	然而，教职人员面临重大挑战。许多在人工智能时代之前接受培训的学者缺乏使用学生每天都在使用的工具的个人经验。学术职业的激励结构——在大多数体系中奖励研究发表甚于教学创新——对投入时间进行教学变革提供的动力有限。而技术变革的速度制造了不确定性：哪些创新值得采纳，哪些只是短暂的潮流。
A study of regulations, technology policies, and university attitudes to AI in China, based on policy analysis and 33 faculty interviews at Chinese research universities, found that faculty generally view AI as enhancing personalization and research productivity but express concerns about academic integrity, algorithmic bias, and over-reliance on automated systems (Higher Education Quarterly 2025). These concerns mirror those of European faculty, suggesting that the faculty experience of AI-driven transformation transcends cultural and systemic differences.	一项基于政策分析和33名中国研究型大学教职人员访谈的关于中国高校人工智能法规、技术政策和态度的研究发现，教职人员普遍认为人工智能增强了个性化和研究生产力，但对学术诚信、算法偏见和过度依赖自动化系统表示担忧（Higher Education Quarterly 2025）。这些担忧与欧洲教职人员的忧虑如出一辙，表明教职人员对人工智能驱动转型的体验超越了文化和制度差异。
6.2 Professional Development for the AI Era	6.2 人工智能时代的教师专业发展
Both European and Chinese universities face the challenge of faculty professional development for AI-augmented teaching. China‘s approach has been characteristically large-scale: the national teacher training platform, integrated into the Smart Education Platform, has trained over 10 million teachers in digital competencies. However, training for AI integration in university teaching specifically remains uneven, with elite research universities moving faster than regional teaching-oriented institutions.	欧洲和中国的大学都面临着人工智能增强教学方面的教师专业发展挑战。中国的方法一如既往地大规模推进：整合到智慧教育平台中的国家教师培训平台已培训了超过1,000万名教师的数字能力。然而，大学教学中人工智能整合的专项培训仍然参差不齐，精英研究型大学的推进速度快于地方教学型院校。
European approaches to faculty development vary by institution but are increasingly supported by EU funding. The Erasmus+ programme’s 2026 call allocates EUR 145.6 million for cooperation partnerships and forward-looking projects, with digital transformation in education as a priority theme. The European University Alliances have also created inter-institutional faculty exchange programmes focused on digital pedagogy.	欧洲的教师发展方法因机构而异，但日益得到欧盟资金的支持。Erasmus+项目2026年度的合作伙伴关系和前瞻性项目征集分配了1.456亿欧元，教育中的数字化转型为优先主题。欧洲大学联盟还创建了以数字教学法为重点的跨机构教师交流项目。
The most effective faculty development programmes share several characteristics: they are discipline-specific rather than generic, they involve hands-on experimentation rather than passive instruction, they create peer communities of practice rather than one-off workshops, and they explicitly address the ethical and pedagogical questions raised by AI rather than focusing solely on technical skills.	最有效的教师发展项目具有若干共同特征：它们是学科特定的而非通用的；它们涉及动手实验而非被动接受指导；它们创建同行实践社群而非一次性工作坊；并且它们明确回应人工智能引发的伦理和教学问题，而非仅聚焦于技术技能。
7. EU vs. China: Two Visions of the Future University	7. 欧盟与中国：未来大学的两种愿景
7.1 The Chinese Model: State-Led Transformation	7.1 中国模式：国家主导的转型
China‘s vision of the future university is characterized by centralized planning, massive digital infrastructure, rapid deployment, and tight integration with national development goals. The State Council’s „Education Modernization 2035“ plan positions education as a strategic national priority and explicitly links educational transformation with technological innovation and economic competitiveness. The „New Generation Artificial Intelligence Development Plan“ (2017) further positions AI competence as a national strategic asset, with implications for every level of education.	中国的未来大学愿景以集中规划、大规模数字基础设施、快速部署和与国家发展目标的紧密整合为特征。国务院"教育现代化2035"方案将教育定位为国家战略优先事项，并明确将教育转型与技术创新和经济竞争力挂钩。"新一代人工智能发展规划"（2017年）进一步将人工智能能力定位为国家战略资产，对各级教育产生深远影响。
The practical manifestation of this vision is impressive in scale. From September 2025, AI education became compulsory in all primary and secondary schools — primary schools focus on AI literacy and exposure, junior high schools on logic and critical thinking, and senior high schools on applied innovation and algorithm design. At the university level, China‘s „Double First-Class“ university initiative (双一流) has created a tier of elite institutions that serve as laboratories for AI-enhanced pedagogy. Peking University, Tsinghua University, and Zhejiang University, among others, have established AI-specific research centres and integrated AI tools into teaching across disciplines.	这一愿景的实践体现在规模上令人瞩目。从2025年9月起，人工智能教育在所有中小学成为必修课——小学侧重人工智能素养和体验，初中侧重逻辑和批判性思维，高中侧重应用创新和算法设计。在大学层面，中国的"双一流"大学倡议创建了一级精英机构，作为人工智能增强教学法的实验室。北京大学、清华大学和浙江大学等建立了人工智能专项研究中心，并在跨学科教学中整合了人工智能工具。
The advantages of this model are evident in scale and speed. No other educational system can deploy a platform serving 293 million learners within two years of launch, or mandate AI education in all primary and secondary schools within a single policy cycle. The near-universal broadband connectivity rate in Chinese schools (99.9 percent of schools connected at 100 Mbps or faster, per the Ministry of Education 2023) (achieved through the Education Informatization 2.0 Action Plan) provides a technological foundation that many European countries have not yet matched.	这种模式的优势在规模和速度上显而易见。没有其他教育体系能在平台推出两年内部署服务2.93亿学习者的平台，或在单个政策周期内强制所有中小学开展人工智能教育。中国学校近乎普及的宽带连接率（根据教育部2023年数据，99.9%的学校接入100 Mbps或更快的宽带，通过教育信息化2.0行动计划实现），提供了许多欧洲国家尚未达到的技术基础。
The disadvantages are equally evident in the constraints on institutional autonomy, the limited space for faculty-driven pedagogical innovation outside state-defined parameters, and the restrictions on access to global AI tools that may limit students’ preparation for international careers. The „walled garden“ approach — promoting domestic AI platforms while restricting access to ChatGPT, Claude, and Gemini — creates a paradox: students are trained in AI literacy using a limited subset of the tools available globally, potentially disadvantaging them in international academic and professional contexts.	劣势同样显而易见——对机构自治的限制、国家设定参数之外教师驱动教学创新的空间有限，以及对全球人工智能工具访问的限制可能影响学生的国际职业准备。"围墙花园"模式——推广国内人工智能平台同时限制访问ChatGPT、Claude和Gemini——制造了一个悖论：学生使用全球可用工具的有限子集接受人工智能素养培训，这可能在国际学术和职业场景中使他们处于不利地位。
7.2 The European Model: Democratic Governance and Humanistic Values	7.2 欧洲模式：民主治理与人文主义价值观
The European vision of the future university is shaped by the tradition of institutional autonomy enshrined in the Magna Charta Universitatum (1988, reaffirmed 2020), the Bologna Process framework, and the European Higher Education Area. The European Universities Initiative, launched in 2019 and significantly expanded since, represents the most ambitious attempt to create a continental university system while preserving institutional diversity. With EUR 145.6 million allocated in the 2026 Erasmus+ call for cooperation partnerships and forward-looking projects, the EU is investing substantially — though not at the scale of China‘s centralized funding.	欧洲的未来大学愿景由《大学大宪章》（1988年签署，2020年重申）所载的机构自治传统、博洛尼亚进程框架和欧洲高等教育区所塑造。欧洲大学倡议于2019年启动并此后大幅扩展，代表了在保留机构多样性的同时创建大陆性大学体系的最雄心勃勃的尝试。2026年Erasmus+合作伙伴关系和前瞻性项目征集中分配了1.456亿欧元，欧盟正在进行大量投资——尽管其规模不及中国的集中化资金。
The intellectual foundation of the European model draws on the concept of Bildung — the German educational ideal of personal formation through engagement with knowledge, culture, and critical inquiry — and its analogues in other European traditions (the French tradition of culture générale, the British tradition of liberal education). These traditions share an emphasis on education as more than skill acquisition: it is the development of the capacity for independent thought, ethical reasoning, and civic participation.	欧洲模式的知识基础借鉴了教养（Bildung）概念——通过参与知识、文化和批判性探究实现个人形成的德国教育理想——以及其他欧洲传统中的类似概念（法国的通识教育传统、英国的博雅教育传统）。这些传统共同强调教育不仅仅是技能习得：它是培养独立思考、伦理推理和公民参与能力的过程。
The EU AI Act provides a distinctive regulatory framework that shapes how European universities approach AI adoption. The Act classifies as „high-risk“ AI systems used to evaluate learning outcomes, determine access to education, and monitor student behaviour during examinations. Emotion recognition in educational settings is banned. These regulatory constraints, which have no Chinese equivalent, create a framework of ethical guardrails that encourages careful, reflective AI integration.	欧盟《人工智能法》提供了独特的监管框架，影响着欧洲大学对人工智能采纳的方式。该法案将用于评估学习成果、确定教育准入和在考试期间监控学生行为的人工智能系统归类为"高风险"。在教育场景中禁止情绪识别。这些在中国没有对应的监管约束，创建了一个伦理护栏框架，鼓励审慎、反思性的人工智能整合。
The advantages of this model lie in its protection of academic freedom, its respect for disciplinary diversity, and its integration of ethical reflection into technological adoption. The disadvantages are fragmentation, slow adoption, uneven implementation across member states, and the risk that democratic governance processes become obstacles to timely adaptation. A comparison of digital transformation across European higher education institutions found significant „similarities and differences“ — the similarities driven by shared challenges and EU policy incentives, the differences reflecting divergent national traditions, funding levels, and institutional cultures (EUA 2025).	这种模式的优势在于对学术自由的保护、对学科多样性的尊重以及将伦理反思融入技术采纳。劣势在于碎片化、采纳缓慢、成员国之间实施不均匀，以及民主治理程序可能成为及时调适的障碍。一项对欧洲高等教育机构数字化转型的比较发现了显著的"相似性与差异性"——相似性源于共同挑战和欧盟政策激励，差异性则反映了不同的国家传统、资金水平和机构文化（EUA 2025）。
7.3 Towards a Synthesis	7.3 走向综合
The most promising model for the university of the future draws on the strengths of both approaches while mitigating their weaknesses. From the Chinese model, it takes the willingness to invest in digital infrastructure at scale, the integration of AI competencies across all disciplines (not only computer science), and the urgency of rapid adaptation. From the European model, it takes the commitment to institutional autonomy, the integration of ethical frameworks into technological adoption, the emphasis on critical thinking and humanistic education alongside technical skills, and the principle that educational transformation should be faculty-led rather than administratively imposed.	最有前景的未来大学模式汲取了两种方法的优势，同时减轻其弱点。从中国模式中借鉴：大规模投资数字基础设施的意愿、在所有学科（而非仅限计算机科学）中整合人工智能能力，以及快速适应的紧迫感。从欧洲模式中借鉴：对机构自治的承诺、将伦理框架融入技术采纳、强调批判性思维和人文教育以及技术技能，以及教育变革应由教师主导而非行政强加的原则。
Concretely, this synthesis might involve: shared digital infrastructure (as in the Chinese model) with institutional freedom in its pedagogical application (as in the European model); mandatory AI literacy across disciplines (as in China‘s new primary-secondary mandate) combined with ethical reflection on AI’s limitations and risks (as in the EU AI Act framework); large-scale data platforms for learning analytics (as in China’s Smart Education Platform) governed by strict data protection principles (as in the GDPR); and national investment in smart campus infrastructure (as in China’s provincial initiatives) deployed through competitive, merit-based processes (as in the Erasmus+ model).	具体而言，这种综合可能包括：共享的数字基础设施（如中国模式）加上教学应用中的机构自由（如欧洲模式）；跨学科的强制人工智能素养（如中国中小学新规）结合对人工智能局限性和风险的伦理反思（如欧盟《人工智能法》框架）；大规模学习分析数据平台（如中国智慧教育平台）受严格数据保护原则约束（如GDPR）；以及国家层面的智慧校园基础设施投资（如中国省级倡议）通过竞争性、绩效导向的程序部署（如Erasmus+模式）。
7.4 Comparative Summary	7.4 比较总结
The following table summarizes the key dimensions of difference between the Chinese and European approaches to the future university:	以下总结了中国和欧洲未来大学方法的关键维度差异：
Infrastructure: China operates a centralized national platform (Smart Education Platform, 293M users); the EU operates decentralized institutional and alliance-based systems (65 alliances, 340 HEIs).	基础设施：中国运营集中式国家平台（智慧教育平台，2.93亿用户）；欧盟运营分散式机构和联盟系统（65个联盟，340所高等教育机构）。
AI Integration: China mandates AI education at all levels from September 2025, with domestic AI tools promoted; the EU encourages AI literacy through the AI Act (February 2025) and institutional policies, with open access to global tools.	人工智能整合：中国自2025年9月起在各级学校强制推行人工智能教育，并推广国内人工智能工具；欧盟通过《人工智能法》（2025年2月）和机构政策鼓励人工智能素养，开放访问全球工具。
Governance: China follows a top-down, state-directed model with the Ministry of Education and State Council setting direction; the EU follows a bottom-up, faculty-governed model with institutional autonomy protected by the Bologna Process.	治理：中国实行自上而下、国家指导的模式，由教育部和国务院设定方向；欧盟实行自下而上、教师治理的模式，机构自治受博洛尼亚进程保护。
Data Environment: China‘s Personal Information Protection Law (PIPL) permits broader educational data use with state oversight; the EU’s GDPR imposes strict data protection requirements that constrain learning analytics.	数据环境：中国的《个人信息保护法》（PIPL）在国家监督下允许更广泛的教育数据使用；欧盟的GDPR施加严格的数据保护要求，制约学习分析。
Speed of Adoption: China‘s centralized model enables rapid deployment across hundreds of millions of users; the EU’s democratic processes ensure careful deliberation but slow implementation.	采纳速度：中国的集中化模式能够在数亿用户中实现快速部署；欧盟的民主程序确保审慎审议但实施缓慢。
Quality Assurance: China relies on institutional accreditation and state-led quality monitoring; the EU relies on disciplinary peer review, institutional accreditation, and the European Standards and Guidelines (ESG).	质量保障：中国依赖机构认证和国家主导的质量监测；欧盟依赖学科同行评审、机构认证和《欧洲标准与准则》（ESG）。
Cultural Foundation: China draws on Confucian educational traditions emphasizing diligence, respect for teachers, and state service; Europe draws on Enlightenment traditions emphasizing critical thinking, individual autonomy, and civic participation.	文化基础：中国借鉴儒家教育传统，强调勤勉、尊师和为国服务；欧洲借鉴启蒙传统，强调批判性思维、个人自主和公民参与。
Neither model fully addresses the core challenge of the future university: integrating technological capability with humanistic purpose. China demonstrates what technology can achieve at scale; Europe demonstrates why scale alone is insufficient without ethical reflection and individual flourishing as educational goals.	两种模式都未完全应对未来大学的核心挑战：将技术能力与人文目的相整合。中国展示了技术在大规模应用中的可能性；欧洲则展示了为什么仅有规模是不够的——缺少伦理反思和个人成长作为教育目标。
7.5 The Skills Question: What Should the Future University Teach?	7.5 技能问题：未来大学应教什么？
Behind the institutional and technological questions lies a more fundamental one: what should the university of the future teach? The traditional answer — disciplinary knowledge organized into degree programmes — is increasingly insufficient in a world where knowledge is abundant, freely accessible, and rapidly obsolescing. The World Economic Forum’s Future of Jobs Report (2024) estimates that 44 percent of workers’ core skills will be disrupted within five years, rendering static skill sets obsolete faster than universities can update their curricula.	在机构和技术问题之下，潜藏着一个更根本的问题：未来大学应教什么？传统答案——以学位项目组织的学科知识——在一个知识丰富、自由可及且快速过时的世界中日益不够充分。世界经济论坛《未来就业报告》（2024）估计，44%的劳动者核心技能将在五年内被颠覆，静态技能组合的过时速度快于大学更新课程的速度。
China has responded to this challenge with characteristic directness, mandating AI literacy as a core competency from primary school onward and integrating „innovation and entrepreneurship“ (创新创业) courses into university curricula across all disciplines. The companion chapter on alternative learning forms (Woesler, this volume) documents how Chinese institutions are also embracing micro-credentials, competency-based education, and project-based learning as supplements to traditional degree programmes.	中国以其一贯的直接方式应对这一挑战，将人工智能素养作为从小学起的核心能力予以强制推行，并在大学各学科课程中融入"创新创业"课程。本论文集中关于替代学习形式的配套章节（Woesler, 本卷）记录了中国机构如何同时拥抱微证书、能力导向教育和项目式学习作为传统学位项目的补充。
European responses have been more varied but increasingly converge on a framework of „transversal competencies“ — critical thinking, creativity, collaboration, communication, digital literacy, and ethical reasoning. The EU’s Key Competences for Lifelong Learning framework (updated 2018) provides a reference point, but implementation at the institutional level remains inconsistent.	欧洲的应对更为多样，但日益趋于"横向能力"框架——批判性思维、创造力、协作、沟通、数字素养和伦理推理。欧盟《终身学习关键能力》框架（2018年更新）提供了参考点，但在机构层面的落实仍然参差不齐。
The deeper question is whether the future university should primarily teach students what to think (knowledge transmission), how to think (critical reasoning), or how to learn (adaptive capacity). The Chinese model, with its emphasis on practical skills and national development priorities, tends toward the first and third. The European model, with its emphasis on Bildung and academic freedom, tends toward the second. The university of the future will need all three — and the wisdom to know when each is appropriate.	更深层的问题在于，未来大学是否应主要教授学生想什么（知识传递）、如何思考（批判性推理）还是如何学习（适应能力）。中国模式侧重于实用技能和国家发展优先领域，倾向于第一和第三种。欧洲模式侧重于教养（Bildung）和学术自由，倾向于第二种。未来大学将需要三者兼备——以及判断何时各有所宜的智慧。
This brings us back to the Confucian-Enlightenment convergence noted earlier. Both traditions recognize that education must go beyond mere knowledge transfer to develop the whole person — the Confucian 君子 (junzi, exemplary person) and the Enlightenment ideal of the autonomous, critically thinking citizen. In an age of artificial intelligence, these humanistic ideals are not obsolete; they are more essential than ever. The distinctly human capacities that both traditions cultivate — moral judgement, aesthetic sensitivity, empathy, the ability to navigate ambiguity and complexity — are precisely the capacities that AI cannot replicate and that the labour market of the future will increasingly value.	这将我们带回前文提到的儒家-启蒙思想的会通。两种传统都认识到教育必须超越单纯的知识传递，培养完整的人——儒家的君子（junzi，模范人格）和启蒙理想的自主的、批判性思考的公民。在人工智能时代，这些人文主义理想并非过时；它们比以往任何时候都更为根本。两种传统共同培养的独特人类能力——道德判断、审美敏感、共情、驾驭模糊与复杂的能力——恰恰是人工智能无法复制的、未来劳动力市场将日益珍视的能力。
8. Conclusion	8. 结论
The university of the future will be neither a Chinese smart campus nor a European humanistic seminar — it will need to be both. The challenge of the coming decade is not to choose between technological efficiency and humanistic depth but to integrate them. AI-personalized learning must coexist with the cultivation of intellectual autonomy. Smart classroom infrastructure must serve pedagogical innovation, not merely facilities management. Generative AI must be integrated into curricula in ways that develop students’ critical thinking rather than replacing it. And hybrid learning must be designed not merely for convenience but for genuine educational effectiveness.	未来大学既不是中国的智慧校园，也不是欧洲的人文主义研讨室——它需要两者兼具。未来十年的挑战不是在技术效率和人文深度之间做出选择，而是将它们整合起来。人工智能个性化学习必须与培养智识自主共存。智慧课堂基础设施必须服务于教学创新，而不仅仅是设施管理。生成式人工智能必须以发展学生批判性思维而非取代之的方式整合到课程中。混合学习必须不仅为了便利而设计，更要追求真正的教育成效。
The comparison of European and Chinese approaches to university transformation reveals that neither system has fully solved the puzzle of the future university. China demonstrates what is possible when a state commits massive resources to educational digitalization; Europe demonstrates what is necessary when educational transformation is guided by democratic values and humanistic traditions. The university that will thrive in 2035 — and in 2050 — will be the one that learns from both traditions, combining scale with subtlety, speed with reflection, and innovation with the enduring human values that have sustained the university for nearly a millennium.	欧中大学转型方法的比较揭示，两个体系都没有完全解决未来大学的难题。中国展示了当一个国家将大量资源投入教育数字化时的可能性；欧洲展示了当教育转型受到民主价值观和人文主义传统引导时的必要性。将在2035年——乃至2050年——蓬勃发展的大学，将是那些从两种传统中学习的大学——将规模与精微、速度与反思、创新与近千年来维系大学的持久人类价值观相结合。
Acknowledgments	致谢
This research was supported by the Jean Monnet Centre of Excellence „EU-Studies Centre: Digitalization in Europe and China„ (EUSC-DEC), funded by the European Union under Grant Agreement No. 101126782. Views and opinions expressed are those of the author only and do not necessarily reflect those of the European Union.	本研究得到了让·莫内卓越中心"欧盟研究中心：中欧数字化"（EUSC-DEC）的支持，由欧盟资助协议编号101126782提供资金。本文所表达的观点仅代表作者本人，不一定反映欧盟的立场。
References	参考文献
Computers and Education: Artificial Intelligence. (2025). Generative AI in higher education: A global perspective of institutional adoption policies and guidelines. Elsevier. DOI: 10.1016/j.caeai.2024.100348.	Computers and Education: Artificial Intelligence. (2025). Generative AI in higher education: A global perspective of institutional adoption policies and guidelines. Elsevier. DOI: 10.1016/j.caeai.2024.100348.
EDUCAUSE. (2025). Shaping the future of learning: AI in higher education.	EDUCAUSE. (2025). Shaping the future of learning: AI in higher education.
Elbertsen, L., Kok, R., & Salimi, N. (2025). Designing the future smart campus: Integrating key elements to enhance user experience. Journal of Science and Technology Policy Management, 16(10), 117–137.	Elbertsen, L., Kok, R., & Salimi, N. (2025). Designing the future smart campus: Integrating key elements to enhance user experience. Journal of Science and Technology Policy Management, 16(10), 117–137.
European Universities Association (EUA). (2025). Similarities and differences in the digital transformation of higher education. Global University Associations Forum (GUAF).	European Universities Association (EUA). (2025). Similarities and differences in the digital transformation of higher education. Global University Associations Forum (GUAF).
International Center for Academic Integrity (ICAI). (2024). Academic integrity and generative AI: Student survey data. As cited in: Frontiers in Artificial Intelligence (2024), „Shaping integrity: Why generative artificial intelligence does not have to undermine education.“ DOI: 10.3389/frai.2024.1471224.	International Center for Academic Integrity (ICAI). (2024). Academic integrity and generative AI: Student survey data. As cited in: Frontiers in Artificial Intelligence (2024), "Shaping integrity: Why generative artificial intelligence does not have to undermine education." DOI: 10.3389/frai.2024.1471224.
Liu, X., et al. (2025). Regulations, technology policies and universities’ attitudes to artificial intelligence in China. Higher Education Quarterly, 79(4). DOI: 10.1111/hequ.70055.	Liu, X., et al. (2025). Regulations, technology policies and universities' attitudes to artificial intelligence in China. Higher Education Quarterly, 79(4). DOI: 10.1111/hequ.70055.
OECD. (2023). Digital Education Outlook 2023: Emerging governance of generative AI in education.	OECD. (2023). Digital Education Outlook 2023: Emerging governance of generative AI in education.
Rize Education. (2025). The hybrid college of the future.	Rize Education. (2025). The hybrid college of the future.
Ruano-Borbalan, J.-C. (2025). The transformative impact of artificial intelligence on higher education: A critical reflection on current trends and future directions. International Journal of Chinese Education. SAGE. DOI: 10.1177/2212585X251319364.	Ruano-Borbalan, J.-C. (2025). The transformative impact of artificial intelligence on higher education: A critical reflection on current trends and future directions. International Journal of Chinese Education. SAGE. DOI: 10.1177/2212585X251319364.
UNESCO. (2023). Smart Education Platform of China: Laureate of UNESCO Prize for ICT in Education.	UNESCO. (2023). Smart Education Platform of China: Laureate of UNESCO Prize for ICT in Education.
Wang, X., Xu, X., Zhang, Y., Hao, S., & Jie, W. (2024). Exploring the impact of artificial intelligence application in personalized learning environments: Thematic analysis of undergraduates’ perceptions in China. Humanities and Social Sciences Communications, 11(1), 1644.	Wang, X., Xu, X., Zhang, Y., Hao, S., & Jie, W. (2024). Exploring the impact of artificial intelligence application in personalized learning environments: Thematic analysis of undergraduates' perceptions in China. Humanities and Social Sciences Communications, 11(1), 1644.
Yu, T., Dai, J., & Wang, C. (2023). Adoption of blended learning: Chinese university students’ perspectives. Humanities and Social Sciences Communications. 10(1), 390. Nature.	Woesler, M. (this volume). Ethical frameworks for AI in higher education: Between European regulation and Chinese innovation.
Woesler, M. (this volume). Ethical frameworks for AI in higher education: Between European regulation and Chinese innovation.	Woesler, M. (this volume). Student data protection in the digital university: GDPR and China's PIPL compared.
Woesler, M. (this volume). Student data protection in the digital university: GDPR and China‘s PIPL compared.	Woesler, M. (this volume). Learning a foreign language with and without AI: An empirical comparative study.
Woesler, M. (this volume). Learning a foreign language with and without AI: An empirical comparative study.	World Economic Forum (WEF). (2023). Future of Jobs Report 2023.
World Economic Forum (WEF). (2023). Future of Jobs Report 2023.	Xie, Q., Li, M., & Cheng, F. (2025). Between regulation and accessibility: How Chinese university students navigate global and domestic generative AI. Globalisation, Societies and Education. Taylor & Francis.
Xie, Q., Li, M., & Cheng, F. (2025). Between regulation and accessibility: How Chinese university students navigate global and domestic generative AI. Globalisation, Societies and Education. Taylor & Francis.	Yaqin, A. M. A., Muqoffi, A. K., Rizalmi, S. R., Pratikno, F. A., & Efranto, R. Y. (2025). Hybrid learning in post-pandemic higher education systems: an analysis using SEM and DNN. Cogent Education, 12(1), 2458930. Taylor & Francis.
Yaqin, A. M. A., Muqoffi, A. K., Rizalmi, S. R., Pratikno, F. A., & Efranto, R. Y. (2025). Hybrid learning in post-pandemic higher education systems: an analysis using SEM and DNN. Cogent Education, 12(1), 2458930. Taylor & Francis.	Yu, T., Dai, J., & Wang, C. (2023). Adoption of blended learning: Chinese university students' perspectives. Humanities and Social Sciences Communications. 10(1), 390. Nature.
Zhang, C. (2026). Unlocking academic gains in smart-classroom settings: A moderated-mediation study among Chinese undergraduates. Acta Psychologica. Elsevier.	Zhang, C. (2026). Unlocking academic gains in smart-classroom settings: A moderated-mediation study among Chinese undergraduates. Acta Psychologica. Elsevier.