《上海交通大学关于在教育教学中使用AI的规范》（试行版）

https://www.sjtu.edu.cn/tg/20250304/207682.html

上海交通大学关于发布《上海交通大学关于在教育教学中使用AI的规范》（试行版）的通知( 2025年03月04日 )

上海交通大学关于在教育教学中使用AI的规范（试行版）

SJTU Guidance on Development and Governance of AI in Education and Teaching (Trial Version)

为贯彻落实上海交通大学《“AI+教育教学”行动方案（2024-2026）》，有序推进高等教育领域人工智能的研发、部署与应用活动，增强各利益相关方的人工智能规范意识与行为自觉，完善“AI+HI（人工智能+人类智慧，Artificial Intelligence + Human Intelligence）”改革治理体系与治理机制，统筹高等教育教学领域人工智能的发展与安全，打造人工智能与人类智慧在“师-生-机-环”四元互动教育教学新生态，特制定本规范。

In order to implement the "AI+ Education and Teaching Action Plan" (2024-2026) of Shanghai Jiao Tong University, systematically advance the research, deployment, and application of artificial intelligence (AI) in the field of higher education, enhance the awareness and behavioral self-discipline of all stakeholders regarding AI norms, improve the reform governance system and mechanism of "AI + HI (Artificial Intelligence + Human Intelligence)", coordinate the development and safety of AI in the field of higher education and teaching, and create a new educational and teaching ecosystem featuring the four-way interaction of "teachers-studentsmachines-environment" with AI and HI, these guidance are hereby formulated.

第一章 总则

Chapter I General provisions

第一条  本规范旨在将高等教育教学领域的规范要求融入“AI+教育教学”改革全生命周期，涵盖场景分析、设计开发、应用部署、评估反馈、调整维护、管理优化等各个环节。通过创新评估体系与标准，在各阶段促进高效、可信、公平、普惠，有效防范技术失能、教育分化、算法歧视、隐私泄露及数据安全、信息安全威胁等风险。

Article 1 The guidance aim to integrate the standard requirements of higher education into the whole life cycle of "AI+ education and teaching" reform, covering various processes such as scenario analysis, design and development, application deployment, evaluation and feedback, adjustment and maintenance, management optimization, etc. By innovating evaluation systems and standards, these guidances promote efficiency, credibility, fairness, and inclusiveness in all processes, effectively preventing risks such as technical incompetence, educational differentiation, algorithmic discrimination, privacy breaches, data insecurity, and cyber threats.

第二条  本规范适用于上海交通大学“AI+教育教学”改革中的“师”（教师）、“生”（学生）、“机”（人工智能技术系统）、“环”（教育教学生态环境）各利益相关方，包括教学管理与支持人员，以及校际合作伙伴和社会资源提供者。

Article 2 These guidances apply to all stakeholders in the "AI + Education and Teaching" reform at SJTU, namely "teachers," "students," "machines" (AI technology systems), and the "environment" (educational and teaching ecological environment), including teaching management and assistant staff, inter-university cooperators, and social resource providers.

第三条  在遵守国内外人工智能法律法规的前提下，上海交通大学鼓励在高等教育教学领域积极探索人工智能使用方式，在各场景推动技术赋能、防范技术风险，构建符合新时代需要的知识生产模式与新型教育体系，赋能学生提升综合素养。

Article 3 While complying with domestic and international AI laws and guidances, Shanghai Jiao Tong University encourages active exploration of the use of AI in the field of higher education and teaching, promotes technology empowerment in various scenarios, promotes technology empowerment and risk prevention in various scenarios, constructs knowledge production models and new educational systems that meet the needs of the new era, and empowers students to enhance their comprehensive literacy.

第四条  本规范所指人工智能（AI），是指通过感知环境、获取知识、推导演绎等方法，对人类智慧的模拟、延伸或扩展，包括但不限于判别式、生成式、数据驱动型、知识图谱驱动型人工智能等多种技术路线。本规范所指人类智慧（HI）即理解、运用、创造知识的独特能力。本规范致力于推动“AI+HI”相结合达成良性人机协同的目标愿景。

Article 4 The term "Artificial Intelligence (AI)" in the guidances refers to the simulation, extension, or expansion of human intelligence through methods such as environmental perception, knowledge acquisition, deduction, and reasoning, including but not limited to various technical routes such as discriminative, generative, data-driven, and knowledge graph-driven AI. The term "Human Intelligence (HI)" refers to the unique ability to understand, apply, and create knowledge. These guidances strive to promote the combination of "AI + HI" to achieve the goal of benign human-machine collaboration.

第五条  上海交通大学“AI+教育教学”各方主体应遵循以下基本原则：

（一）技术以人为本、以学为本。围绕“让每个学生更优秀”育人理念，发展应用有利于学生知识能力增长、有利于教师教学育人的人工智能技术。定期评估师生对工具的使用情况，确保技术应用辅助教学目标，避免教学任务的完全替代或片面依赖。

（二）技术合法合规、可靠可信。教育教学改革中发展应用的人工智能技术服务或产品应符合前述法律规范要求，确保公平公正、责任明确、透明可追溯，人工智能应用应落实责任主体和追责流程。

（三）尊重多元精神、推动开放协作。应尊重高等教育学科和知识多样性，结合学科特点，促进知识融合创新，探索自主、自发的个性化学习模式，并关注少数群体或跨学科学生，确保共享教育资源的广泛可及性。

（四）鼓励改革实验、包容探索创新。应充分发挥师生主动性、积极性，鼓励在教育教学各环节、各领域广泛开展人工智能实验，探索不同场景下“AI+HI”的优化方案，累进迭代、包容试错。在遵守学术规范、教学底线与伦理边界的前提下，不得破坏现有教学秩序。

（五）提升伦理素养、鼓励敏捷治理。通过教育教学活动普及人工智能伦理治理知识，敏捷应对可能出现的风险，逐步完善治理规范，应针对不同应用场景设立应急预案，并构建师生快速响应机制。

Article 5 All parties involved in "AI+ Education and teaching" at Shanghai Jiao Tong University shall follow the following basic principles:

(1) Human-Centered and Learning-Focused Technology: Centering on the educational philosophy of "making every student better," develop and apply AI technologies that benefit students' intellectual growth and teachers' instructional and educational efforts. Regularly assess the use situations of tools by teachers' and students', to ensure that technology application assists teaching goals and avoids complete substitution or singlefaceted reliance on teaching tasks.

(2) Legal, Compliant, Reliable, and Trustworthy Technology: AI services or products developed and applied in educational and teaching reforms should comply with the legal and regulatory requirements, ensuring fairness, clear responsibility, transparency, and traceability. AI applications should establish responsible entities and accountability processes.

(3) Respect for Diversity and Promotion of Open Collaboration: Respect the diversity of disciplines and knowledge in higher education, promote knowledge integration and innovation, explore autonomous and spontaneous personalized learning modes based on disciplinary characteristics, and pay attention to minority groups or interdisciplinary students to ensure widespread accessibility to shared educational resources.

(4) Encourage Experimental Reform and Tolerate Exploratory Innovation: Fully leverage the initiative and enthusiasm of teachers and students, encourage extensive AI experiments in all aspects and fields of education and teaching, explore optimization schemes for "AI + HI" in different scenarios, iteratively improve with tolerance for errors, and do not disrupt the existing teaching order while adhering to academic norms, teaching bottom lines, and ethical boundaries.

(5) Enhance Ethical Literacy and Encourage Agile Governance: Popularize knowledge on AI ethical governance through educational and teaching activities, agilely respond to potential risks, gradually improve governance norms, establish emergency response plans for different application scenarios, and construct a rapid response mechanism for teachers and students.

第六条  根据应用风险的差异，本规范将高等教育教学领域的人工智能应用划分为四种类型：禁止使用、有限使用、鼓励使用、开放使用，从而推进“AI+教育教学”分级分类改革。

1  禁止使用

指存在高风险，或者师生经充分讨论一致反对的场景或环节。禁止使用的场景应明确列入学校制定的“人工智能应用禁止清单”，通过校内平台及时更新、透明公开，保障师生了解相应信息并遵守。

2  有限使用

指存在一定风险，需结合风险评估和师生讨论意见，在特定条件下限制使用的场景或环节。有限使用应根据以下参考指标进行动态评估：

隐私影响：对师生个人信息保护的潜在威胁；

学习成效影响：对教学目标和学习效果可能造成的负面影响；

安全合规性：是否符合相关法律法规及学校规范要求。

对有限使用场景，应设置具体的约束条件，并定期开展再评估。

3  鼓励使用

指风险较低，或师生经讨论达成较高共识，推动积极应用的场景或环节。鼓励使用的人工智能工具或技术需通过学校备案流程，确保其符合教育教学目标、伦理规范和技术可靠性要求。

4  开放使用

指趋近零风险，或师生经讨论一致认为适宜广泛使用的场景或环节。开放使用的人工智能技术应具备较高的普适性和公平性，能够广泛惠及师生，并通过学校平台向全校推广。

此外，对于师生经充分讨论后仍存在分歧的场景或环节，可引入仲裁或复议机制。由学校人工智能教育教学改革治理委员会牵头，结合技术评估、风险分析和教育目标对争议进行仲裁，形成最终清单。

Article 6 According to the differences in application risks, these guidances classify AI applications in the field of higher education and teaching into four types: Prohibited Use, Limited Use, Encouraged Use, and Open Use, thereby promoting a multi-level reform of "AI + Education and Teaching."

1. Prohibited Use: Refers to scenarios or links with high risks or that teachers and students have unanimously opposed to upon thorough discussion. Scenarios prohibited from use should be clearly listed in the "AI Application Prohibition List" of SJTU, updated timely through the internal platform, and publicly disclosed to ensure that teachers and students are informed and comply accordingly.

2. Limited Use: Refers to scenarios or links with certain risks that require restricted use under specific conditions based on risk assessments as well as teacher-student discussions. Limited use should be dynamically assessed based on the following reference indicators:

 Privacy impact: Potential threat to the protection of personal information of teachers and students;

 Impact on learning effectiveness: Possible negative impact on teaching objectives and learning outcomes;

 Safety compliance: Whether it complies with relevant laws, guidances, and norms of the university.

Specific constraints should be set for limited-use scenarios, and reassessments should be conducted regularly.

3. Encouraged Use: Refers to scenarios or links with low risks or that teachers and students have reached a high consensus through discussion to promote application. AI tools or technologies encouraged for use need to pass the evaluation process of SJTU to ensure they meet educational and teaching goals, ethical norms, and technical reliability requirements.

4. Open Use: Refers to scenarios or cases with near-zero risks or that teachers and students have unanimously agreed on through discussion for wide use. AI technologies for open use should guarantee fairness, widely benefit teachers and students, and be promoted through the platforms of SJTU.

Furthermore, for scenarios or processes where disagreements persist between teachers and students after thorough discussion, an arbitration or review process can be introduced. The arbitration will be made based on technical evaluations, risk analyses, and educational objectives by the AI Education and Teaching Reform Governance Committee. Then, the committee will finalize the use list.

第七条  高等教育教学领域人工智能应用的规范使用要求，涵盖事前、事中、事后等环节，具体包括承诺、备案、影响评估、风险记录与管理、透明与可解释、利益相关方咨询等机制措施。针对不同类型的人工智能应用，高等教育教学领域各场景自行制定、完善与之相匹配的规范使用要求。

Article 7 The requirements for the standardized use of AI applications in the field of higher education and teaching cover different processes of before, during and after, including commitment, filing, impact assessment, risk recording and management, transparency and explainability, stakeholder consultation and other mechanisms and measures. For different types of AI applications, each scenario in the field of higher education and teaching should develop and improve the corresponding standard use requirements.

第二章 教师规范

Chapter II Guidances on Teachers

第八条  教师应积极探索人工智能赋能教育的可能性，明确人工智能辅助教学授课的原则。根据“禁止使用”“有限使用”“鼓励使用”“开放使用”原则，在教学材料生产、个性化教学、学习进程跟踪与评估、师生互动等场景中，合理使用人工智能技术，有效提升教学质量。

Article 8 Teachers should actively explore the possibilities of AI empowering education and clarify the principles of AI-assisted teaching. Based on the principles of "Prohibited Use," "Limited Use," "Encouraged Use," and "Open Use," teachers should reasonably use AI technologies in scenarios such as teaching material preparation, personalized teaching, learning progress tracking and assessment, and teacher-student interaction to effectively improve teaching quality.

第九条 教师是AI+教学设计的第一责任人，应在遵守教育教学相关规章制度要求的前提下，合理、合规、有效使用人工智能技术产品或服务，必要时提供人工智能测试验证报告及备案，包括平台选择、使用范围及方式，并围绕场景应用、风险事项提示、应急预案等各环节设置具体规范。对于使用未备案或存在风险的人工智能工具情况，学校可追究相关责任。

Article 9 Teachers are the primary responsible persons for AI + teaching design. Teachers should reasonably, compliantly, and effectively use AI technology products or services while complying with relevant rules and guidances on education and teaching, and provide AI test and validation reports and filings if necessary, including platform selection, scope and method of use, and set specific norms for scenario application, risk reminders, and emergency response plans. SJTU may pursue the relevant responsibilities to those who use unregistered or risky AI tools.

第十条  教师应充分关注学生在人工智能辅助场景中的自主性与获得感。通过教学反馈机制、学习成效评估等手段，助力学生平衡自主学习与人工智能支持，并关注学生差异化需求，避免因技术门槛或资源不均带来的负面效应。

Article 10 Teachers should fully pay attention to students' autonomy and sense of achievement in AI-assisted scenarios. Through teaching feedback mechanisms and learning effectiveness assessments, teachers should help students balance autonomous learning and AI support, pay attention to students' differentiated needs, and avoid negative effects caused by technical thresholds or uneven resource distribution.

第十一条  教师应立足各项教育教学场景，有效治理价值失范现象。在教育教学过程中与学生等主体充分讨论，通过教学大纲、课堂规则、评分标准等形式，制定学生使用人工智能工具的细则要求，对于人工智能工具使用范围、人工智能生成信息引用模板等予以公开说明。

Article 11 Teachers should effectively govern value anomie phenomena based on various educational and teaching scenarios. During the educational and teaching process, teachers should fully discuss with students and other entities, formulate detailed requirements for students' use of AI tools through teaching syllabus, classroom rules, scoring standards, etc., and publicly explain the scope of AI tool use and templates for citing AI-generated information.

第三章 学生规范

Chapter III Guidances on Students

第十二条  学生应结合课堂要求和指导方案，合理选择人工智能工具，了解人工智能“禁止使用”“有限使用”“鼓励使用”“开放使用”的具体场景，构建良性人机协同的新常态。

Article 12 Students should reasonably choose AI tools based on teaching requirements and guidance, understand the specific scenarios of "Prohibited Use," "Limited Use," "Encouraged Use," and "Open Use" of AI, and establish a new normal for benign human-machine collaboration.

第十三条  学生应了解并遵守各项课程的人工智能使用规范，在课堂学习、作业反馈等环节，遵循教学计划、知识产权等法律法规、学术诚信要求。若有违规行为，学校将追究相关责任。

Article 13 Students should understand and comply with the AI use guidances for each course, and follow teaching plans, laws and guidances such as intellectual property, and academic integrity requirements in classroom learning, homework feedback, and other study processes. The university will hold the relevant individuals accountable.

第十四条  学生应积极提升人工智能素养，利用学校提供的人工智能课程及学习资源，通过“AI学习互助小组”等合作网络搭建，识别人工智能相对人类智慧的优势劣势，理性认知人工智能的局限。

Article 14 Students should actively enhance their AI literacy by utilizing the AI courses and learning resources provided by the university. Through collaborative networks such as "AI Learning Mutual Help Groups," they should identify the strengths and weaknesses of AI compared to human intelligence and develop a rational understanding of AI's limitations.

第十五条  学生应主动参与高等教育领域的人工智能创新改革，自主参与人工智能学术研讨与公益活动，探索多学科“AI+”范式，为校内人工智能创新应用积极献策。

Article 15 Students should actively participate in AI innovation reforms in the field of higher education, autonomously engage in AI academic discussions and public welfare activities, explore multidisciplinary "AI+" paradigms, and actively contribute ideas for AI innovation applications at SJTU.

第十六条  学生应坚持人工智能辅助学习的价值定位，形成自适应的学习模式，自觉标注与引用人工智能工具的必要信息，在知识学习与创新创造的人机协同过程中培养自主能力、反思能力，以人工智能促进人类智慧的发展。

Article 16 Students should adhere to the value orientation of using AI as an auxiliary learning tool, forming an adaptive learning mode, and consciously annotating and citing necessary information about AI tools. In the process of knowledge acquisition and innovative creation in human-AI collaboration, students should cultivate their autonomy and reflective abilities, promoting the development of human intelligence through AI.

第四章 AI产品和服务规范

Chapter IV Guidances on AI Products and Services

第十七条  人工智能技术产品和服务开发流程不得用于违背学术诚信、窃取信息、侵犯隐私的行为。其开发场景须符合人工智能“禁止使用”“有限使用”清单规范。

Article 17 The development process of AI technology products and services shall not be used for actions that violate academic integrity, steal information, or infringe upon privacy. Their development scenarios must comply with the "Prohibited Use" and "Limited Use" list guidances for AI.

第十八条  人工智能技术产品和服务的开发者应提供算法透明度评估报告，说明核心算法逻辑、数据来源及模型训练方法，确保其可解释性。

Article 18 Developers of AI technology products and services should provide algorithm transparency assessment reports, explaining the core algorithm logic, data sources, and model training methods to ensure their interpretability.

第十九条  人工智能技术产品或服务应遵循可追溯性原则，对于借助人工智能产生的信息集合推行溯源、标识要求。在涉及学校相关公共数据时，使用人工智能产品或服务须经学校批准或备案。

Article 19 AI technology products or services should follow the principle of traceability, implementing tracing and identification requirements for information sets generated with the assistance of AI. When AI products or services involve the university's related public data, their use must be approved or registered with the university.

第五章 利益相关方规范

Chapter V Guidances on Stakeholders

第二十条  学校成立人工智能教育教学改革治理委员会，涵盖“师-生-机-环”各领域代表，明确各方在技术评估、伦理审查、师生反馈整合等方面的职责分工，确保治理委员会高效运作。在各一线教育教学单位成立治理小组，探索创新案例、总结治理经验、制定针对具体场景的规范要求。针对人工智能应用过程中的争议问题，鼓励展开“师-生-机-环”大讨论，在相互评议中促进集体共识。

Article 20 The university establishes an AI Education and Teaching Reform Governance Committee, which covers representatives from various fields including "teachers, students, machines, and environment". The different responsibilities among all parties in technical assessment, ethical review, and integration of teacher and student feedback will be clarified to ensure efficient operations of the governance committee. Governance groups should be established in schools or institutes of SJTU to explore innovative cases, summarize governance experiences, and formulate regulatory requirements for specific scenarios. Regarding controversial issues arising from AI applications, it is encouraged to organize "teacher-student-machine-environment" discussions to promote collective consensus through mutual comments.

第二十一条  学校职能部门应不断完善“AI+教育教学”改革治理框架，完善高等教育场景下人工智能技术系统及应用的风险评估和分级分类管理体系。应鼓励一线教育教学单位开展高等教育领域的人工智能社会实验，探索人工智能的多用途、多场景应用标杆案例。

Article 21 All the management departments of SJTU should continuously improve the reform governance framework for "AI + Education and Teaching," perfecting the risk assessment and hierarchical and categorical management system for AI technology systems and applications in higher education scenarios. All the education and teaching units should be encouraged to conduct social experiments on AI in higher education and to explore benchmark cases of AI's multi-purpose and multi-scenario applications.

第二十二条  鼓励科技企业及其他市场主体积极参与“AI+教育教学”改革，提供合乎治理规范要求的人工智能技术产品与服务。应探索多元治理路线、敏捷回应师生反馈，形成合作、信任、透明的协同建设新生态。

Article 22 Technology enterprises and other market entities are encouraged to actively participate in the "AI + Education and Teaching" reform by providing AI technology products and services that meet governance regulatory requirements. They should explore multiple governance routes, agilely respond to teacher and student feedback, and form a new ecosystem of cooperation, trust, and transparency.

第二十三条  人工智能技术产品或服务的研究开发者、设计制造者、部署应用者应强化责任意识，充分认识、全面分析高等教育领域人工智能应用的伦理治理安全风险，按照分级分类、激励相容原则建立必要保障机制，对相应引致损失提供救济。

Article 23 Researchers, developers, designers, manufacturers, and deployers of AI technology products or services should strengthen their sense of responsibility and fully recognize and comprehensively analyze the ethical governance and security risks associated with AI applications in higher education. They should establish necessary safeguard mechanisms based on hierarchical classification and incentive compatibility principles, providing relief for losses incurred accordingly.

第二十四条  第三方社会力量是“AI+教育教学”生态的重要参与者。应鼓励技术社群、社会组织、咨询机构等以不同形式参与高等教育领域的人工智能应用改革，提供多方意见、形成社会监督、共促可靠可信。

Article 24 Third-party social forces are important participants in the "AI + Education and Teaching" ecosystem. Technical communities, social organizations, and consulting institutions should be encouraged to participate in AI application reform in higher education in various forms, provide multiple perspectives, form social supervision, and to jointly promote reliability and trustworthiness.

第六章 附则

Chapter VI Supplementary Provisions

第二十五条  本规范由上海交通大学发布，并负责解释和指导实施。同时本规范将根据人工智能技术发展，以及“AI+教育教学”改革进程适时修订。

Article 25 The guidances are issued by Shanghai Jiao Tong University, which is responsible for interpreting and guiding their implementation. Meanwhile, the guidances will be revised in a timely manner according to the development of AI technology and the progress of the "AI + Education and Teaching" reform.