Harvard

HGSE Policy on Student Use of Generative Artificial Intelligence in Academic Work

• 官方来源Official source https://registrar.gse.harvard.edu/learning/policies-forms/ai-policy
• 域名Domain registrar.gse.harvard.edu

Generative Artificial Intelligence (AI) poses both great opportunities and great challenges for the field of education. Tools such as ChatGPT, DALL-E, and GitHub Copilot are having a profound influence on teaching and learning – and on your role as education practitioners and leaders. Your time at HGSE is an opportunity to learn to leverage such tools to produce more equitable access to and deeper engagement in education.

HGSE encourages responsible experimentation with generative AI tools, but there are important considerations to keep in mind when using these tools, including information security and data privacy, copyright issues, the trustworthiness of the content they generate, and academic integrity. The implications of using generative AI for your own learning are equally salient. It might be possible to use this technology to complete some class assignments while doing little work yourself, but short cutting the process of thinking and writing in this way would rob you of the learning you came to HGSE to experience. At its best, generative AI can be like a tutor or thought partner with unlimited time to help you learn – but it should not be used to do the cognitive work for you, or else your own learning will be greatly diminished.^{学习理由Learning rationale摘录Excerpt}

The following guidelines aim to ensure that, in your academic work at HGSE, you use generative AI when it can help you learn and not when it is a hindrance. If you have any doubt about whether a specific use of generative AI is permitted for an assignment or course, you are responsible for discussing it with your instructor prior to using it.

1. Unless otherwise specified by your instructor, it is a violation of the HGSE Academic Integrity Policy to use generative AI to create all or part of an assignment for a course (e.g., a paper, memo, presentation, or short response) and submit it as your own.^{默认规则Default rule} This rule parallels other rules.^{违规定性Integrity framing} You may not ask another person to complete your assignment for a course. You may not copy from something someone else has created or re-write in your own words something someone else has written, without proper attribution.
2. Permissible uses of generative AI in HGSE coursework include seeking clarification on concepts, brainstorming ideas, or generating scenarios that help contextualize what you are learning.^{分级许可Tiered scale} For instance, it is fine to use AI-powered web search and to have “conversations” with tools like ChatGPT to help you explore ideas, refine your thinking, identify examples, and better understand course material. It is also acceptable to use generative AI to draft emails to instructors, students, and others in the HGSE community that are not being submitted as coursework.
3. For any permitted use of GenAI tools, you must acknowledge and document that use in your assignment submission by explaining what tool(s) you used, prompts you provided (if applicable), and how you integrated the output into your work.^{披露要求Disclosure} If you cite directly from the tool, use proper citation format to credit the source.^{过程留证Process evidence} For more details and examples, see these APA guidelines: https://apastyle.apa.org/blog/how-to-cite-chatgpt
4. Keep in mind that the information provided by generative AI tools like ChatGPT is generated from unverified crowd-sourced information. Large language models can produce false claims or “hallucinations” and will regenerate any biases in the corpus of texts on which they are trained. You therefore should not trust the information as if it were equivalent to published research. You are ultimately responsible for the accuracy of the work you submit.^{核实责任Verification duty}
5. The use of generative AI may also have implications for the protection of your own intellectual property. For example, if you upload your own original content to a generative AI tool, that content may become part of the tool’s models, which others may encounter and use.^{知识产权IP & copyright} Conversely, if you use generative AI to develop your own original work, it may unexpectedly include others' copyrighted material.
6. Certain uses of AI also infringe on copyright laws applicable to U.S. universities or contravene existing expectations for student conduct in HGSE courses. For example, the HGSE Student Handbook notes that students may not post, publish, sell, or otherwise publicly distribute course materials without the written permission of the course instructor. Such materials include lecture slides, video or audio recordings, assignments, problem sets, examinations, other students’ work, and answer keys. Students may not make recordings of course material for their own use without written permission of the instructor.
7. Uploading any substantial course content — including text, video, readings, discussion-board pages, or audio recordings — is only allowable through the Harvard-approved AI Sandbox, which ensures data entered is kept in a secure environment and not used to train public AI tools.^{机构工具Institutional tools} The Sandbox is available through individual courses; if you have questions about the Sandbox, discuss with your instructor.
8. It is forbidden to make your own recording of any course meetings, with or without AI tool integrations. If you require or would prefer that course meetings be recorded, discuss this request with your instructor. More broadly, if you require AI technology as part of an assistive technology solution to enable you to participate fully in the course, you must coordinate your usage with the Office of Student Affairs.
9. Given the wide range of learning goals in courses at HGSE, individual instructors may create course-specific policies that differ from and supersede these guidelines.^{课程规则优先Instructor override} Again, if you have any doubt about whether a specific use of generative AI is permitted for an assignment or course, you are responsible for discussing it with your instructor prior to using it.

New ways of teaching and learning will emerge as generative AI becomes increasingly ubiquitous and robust. We thus anticipate that this policy will also evolve, with feedback from students and instructors.

University-wide Guidelines for Using ChatGPT and Other Generative AI Tools

• 官方来源Official source https://provost.harvard.edu/guidelines-using-chatgpt-and-other-generative-ai-tools-harvard
• 域名Domain provost.harvard.edu

Harvard supports responsible experimentation with generative AI tools, but identifies important considerations including information security and data privacy, compliance, copyright, and academic integrity.

Initial guidelines for use of generative AI tools

• Protect confidential data: Do not enter data classified as confidential (Level 2 and above), including non-public research data, into publicly available generative AI tools. Information shared with generative AI tools using default settings is not private and could expose proprietary or sensitive information to unauthorized parties.^{数据与隐私Data & privacy}
• You are responsible for any content that you produce or publish that includes AI-generated material: AI-generated content can be inaccurate, misleading, or entirely fabricated (“hallucinations”), or may contain copyrighted material. Review AI-generated content before publication.
• Adhere to current policies on academic integrity: Review the relevant School’s student and faculty handbooks and policies. Faculty should be clear with students about permitted uses, if any, of generative AI in classes and academic work. Students are encouraged to ask instructors for clarification as needed.
• Be alert for AI-enabled phishing: Generative AI has made sophisticated scams easier to create at scale. Continue to follow security best practices and report suspicious messages.
• Connect with HUIT before procuring generative AI tools: Tools procured on behalf of Harvard should have appropriate privacy and security protections. Vendor generative AI tools must be assessed for risk by Harvard’s Information Security and Data Privacy office before use.

Harvard notes that these guidelines are not new University policy; rather, they apply existing University policies to generative AI.