Teaching & Learning
- The student handbook explicitly categorizes use of unauthorized materials (including generative AI) on assignments as an Honor Code cheating violation unless expressly permitted by the instructor, with handling and sanctions varying by violation level
- Use of generative AI for coursework and assignments is generally governed by instructor-specific policies; unless an instructor explicitly permits it, students are expected to treat generative AI as an unauthorized resource and its use can constitute an Honor Code violation
Each individual instructor can establish policies applicable to academic work in their courses.
Unless specifically permitted by the instructor, students should assume generative AI is unauthorized.
Use of unauthorized materials, including generative AI, on an assignment worth less than 10% of the final course grade, such as a homework assignment or blog post, unless the use of such materials has been expressly permitted by the instructor.
Use of unauthorized materials, including generative AI, on an assignment worth between 10% and 25% of the final course grade, unless the use of such materials has been expressly permitted by the instructor.
Use of unauthorized materials, including generative AI, on an assignment worth more than 25% of the final course grade, unless the use of such materials has been expressly permitted by the instructor.
- For William & Mary Law School, generative AI is prohibited during a student’s exam by default (though instructors may deviate with written notice)
- In the broader university undergraduate Honor Code examples, using materials or equipment during a quiz or examination without explicit instructor authorization is listed as an Honor Code cheating example; the AI-specific prohibition for exams is otherwise handled via instructor authorization expectations rather than a universal campus-wide exam rule
May not be used during a student’s exam.
Instructors have discretion to deviate from the default rule, provided that they do so in writing and with appropriate notice.
Using or viewing any materials or equipment, including a cell phone, laptop/tablet or other electronic device, during a quiz or examination without explicit instructor authorization
- William & Mary Law School’s default policy permits students to use study materials prepared by generative AI
- Separately, university guidance encourages students to verify AI output accuracy and reliability when using AI as a learning aid
A student may use study materials prepared by Generative AI.
When using AI, we encourage you to consult the Studio for Teaching & Learning Innovation’s Evaluating AI Outputs Using Lateral Reading Strategies. This document provides a helpful checklist to verify the accuracy and reliability of AI output.
- The university’s GenAI guidelines require human review for accuracy and security when generating code with GenAI before posting/uploading to university systems
- ChatGPT Edu is described as supporting activities including writing code, but academic permissibility for graded coding assignments remains governed by instructor authorization expectations rather than a single institution-wide classroom rule
Maintain a Human-in-the-Loop and Adhere to W&M Core Values: Review the output of GenAI applications for potential plagiarism, accuracy and security, and ensure it meets the university’s standards for principles of equity, ethics, and appropriateness. For example, when generating code using GenAI, check for accuracy and security prior to posting/uploading on any university systems.
Whether you're brainstorming ideas, summarizing complex readings, writing code, or exploring new topics, ChatGPT Edu can help you think more deeply, work more efficiently, and collaborate more creatively.
Research
- William & Mary’s GenAI guidelines require transparency for GenAI output used for work purposes, which can apply to research-related writing when community members use GenAI tools
- The AIPI final report identifies risks around computer-generated scholarly text/images without proper attribution and recommends adherence to academic integrity and copyright norms, but it is framed as recommendations rather than a binding, detailed manuscript policy
Maintain Transparency: Clearly attribute any output used for work purposes to the GenAI application that created it.
In the context of research, GenAI enables faculty to produce computer-generated text or images without proper attribution, potentially blurring authorship boundaries and raising questions about the originality and authenticity of their scholarly contributions.
GenAI users should adhere to the university’s code of conduct around academic integrity, plagiarism and the use of unauthorized resources.
- AIPI materials discuss GenAI use for data analysis and predictive modeling in institutional contexts and emphasize caution when using GenAI with student information or other sensitive university data
- The GenAI guidelines prohibit inputting protected/sensitive information or PII into non-approved GenAI applications, which would constrain use of GenAI in research data handling when such data are involved
However, the university community must use caution when leveraging GenAI with student information or other sensitive university data.
Do not input university intellectual property, protected or sensitive information, or personally identifiable information (PII) into non-approved GenAI applications.
- The university’s GenAI guidelines require compliance with applicable laws and university policies and emphasize human review to check for plagiarism, accuracy, and security
- The AIPI final report explicitly ties GenAI use to academic integrity and plagiarism expectations and discusses transparency and attribution as elements of research integrity, but it does not provide a specific institutional rule for AI use in IRB or grant submissions in the cited excerpts
Maintain Compliance: Comply with applicable commonwealth and federal laws, regulations (FERPA, HIPAA, etc.), and university policies regarding the use or development of GenAI content or tools.
Maintain a Human-in-the-Loop and Adhere to W&M Core Values: Review the output of GenAI applications for potential plagiarism, accuracy and security, and ensure it meets the university’s standards for principles of equity, ethics, and appropriateness.
GenAI users should adhere to the university’s code of conduct around academic integrity, plagiarism and the use of unauthorized resources.
Academic Integrity
- For work purposes, William & Mary’s GenAI guidelines require users to attribute GenAI output used
- The provided sources do not indicate a single mandatory, university-wide citation format for AI in student submissions
- For coursework, GenAI use is contingent on instructor permission; if permitted, the Writing Resources Center advises students to cite/acknowledge that use (e.g., include a statement of use at the end of the paper, an appendix recording and reflecting on use, or a similar acknowledgment)
Maintain Transparency: Clearly attribute any output used for work purposes to the GenAI application that created it.
If you are using GenAI and have your professor’s permission, you must cite this usage.
The Writing Resources Center is asking that students who have permission to use GenAI in their writing to include a statement of use at the end of their papers, an appendix in which they record and reflect on their use of the technologies, or a similar acknowledgment.
As you write your paper, we recommend you treat GenAI like a research partner and acknowledge its contributions in a note.
- The university frames unauthorized GenAI use as potentially constituting an Honor Code violation and encourages faculty to report suspected unauthorized use to the Honor Council
- The student handbook provides an enforcement structure via the Honor Code process and examples of sanctions for cheating, explicitly including use of unauthorized materials such as generative AI, but the provided sources do not define a university-wide stance on the use of AI detection tools as evidence
Use of unauthorized resources may be considered an Honor Code violation, and we encourage you to report such matters to the Honor Council.
Use of unauthorized materials, including generative AI, on an assignment worth less than 10% of the final course grade, such as a homework assignment or blog post, unless the use of such materials has been expressly permitted by the instructor.
Use of unauthorized materials, including generative AI, on an assignment worth more than 25% of the final course grade, unless the use of such materials has been expressly permitted by the instructor.
Institutional & Administrative
- Faculty guidance explicitly states that instructors determine what AI resources are permitted for students’ assignments and that the university does not have a single one-size-fits-all classroom AI policy
- The GenAI guidelines apply to faculty and staff and require security-policy compliance, attribution of GenAI outputs used for work purposes, and human review of outputs for plagiarism, accuracy, and security
This guideline applies to William & Mary community members (including faculty, staff, and students) using GenAI applications or the development of GenAI models.
Maintain Transparency: Clearly attribute any output used for work purposes to the GenAI application that created it.
Maintain a Human-in-the-Loop and Adhere to W&M Core Values: Review the output of GenAI applications for potential plagiarism, accuracy and security, and ensure it meets the university’s standards for principles of equity, ethics, and appropriateness.
As with all resources, the individual instructor determines what resources are permitted for use in completing assignments for their courses. The university does not have a one-size-fits-all AI policy.
- Personal/non-Edu ChatGPT accounts are not approved for sensitive or non-public data
- W&M IT maintains an AI tool review process and lists approved tools; it also prohibits access to DeepSeek AI on the university network and university-owned devices effective February 18, 2025
- W&M’s GenAI guidelines prohibit inputting university intellectual property, protected/sensitive information, or PII into non-approved GenAI applications, and instruct users to follow university security policies and use W&M SSO for university-provided GenAI applications
Do not input university intellectual property, protected or sensitive information, or personally identifiable information (PII) into non-approved GenAI applications.
Follow William & Mary Security Policies: Review the university's data and information security policies. Only use W&M Single Sign-on (SSO) for logging into university-provided GenAI applications.
As new AI tools and services are introduced, and new AI features are integrated into existing tools, IT will review these tools and services to confirm they are accessible and secure. This is required even for free tools or services.
Effective February 18, 2025, DeepSeek AI and other applications owned by the Chinese company Hangzhou DeepSeek Artificial Intelligence Basic Technology Research Co., Ltd are prohibited on the university’s network (including VIMS) and university-owned devices.
William & Mary’s agreement ensures that your data is not used to train OpenAI models or to improve their services.
Copilot with commercial data protection is enabled by default for all faculty and staff at W&M and is the recommended version for use by faculty and staff.
When Copilot with commercial data protection is enabled: ...Chat history is not saved, prompts and responses are not used to train the large language models...
Please note, however, that these accounts are not approved for any sensitive or non-public data.
- W&M states it is actively exploring responsible ways to integrate AI across teaching, research, and operations
- The AIPI final report calls for coordination and governance including publishing clear expectations, policies/standards/guidelines and approved AI tools, and it recommends establishing a GenAI innovation and policy committee to update and maintain related policies and guidelines
As artificial intelligence (AI) continues to evolve and shape higher education, William & Mary is actively exploring how these emerging technologies can support its mission in teaching, research, and operations.
Coordination and Governance. The campus community should publish a clear set of expectations and best practices around leveraging GenAI in university operations. This includes developing and communicating policies, standards, guidelines, and approved AI tools. This also includes a GenAI innovation and policy committee to update and maintain related policies and guidelines.
DocuMark: Responsible AI Use for Academic Integrity
Knowing your institution's AI policy is step one. DocuMark helps enforce it fairly by empowering universities to manage AI-generated content, prevent cheating, and support student writing through responsible AI use.
Common Questions About College of William & Mary's AI Policies
College of William & Mary has defined AI policies in 12 of 12 categories, with an overall coverage score of 100%.
For work purposes, William & Mary’s GenAI guidelines require users to attribute GenAI output used. For coursework, GenAI use is contingent on instructor permission; if permitted, the Writing Resources Center advises students to cite/acknowledge that use (e.g., include a statement of use at the end of the paper, an appendix recording and reflecting on use, or a similar acknowledgment). The provided sources do not indicate a single mandatory, university-wide citation format for AI in student submissions.
The university frames unauthorized GenAI use as potentially constituting an Honor Code violation and encourages faculty to report suspected unauthorized use to the Honor Council. The student handbook provides an enforcement structure via the Honor Code process and examples of sanctions for cheating, explicitly including use of unauthorized materials such as generative AI, but the provided sources do not define a university-wide stance on the use of AI detection tools as evidence.
W&M’s GenAI guidelines prohibit inputting university intellectual property, protected/sensitive information, or PII into non-approved GenAI applications, and instruct users to follow university security policies and use W&M SSO for university-provided GenAI applications. W&M IT maintains an AI tool review process and lists approved tools; it also prohibits access to DeepSeek AI on the university network and university-owned devices effective February 18, 2025. Approved tools include ChatGPT Edu (under an agreement that W&M data is not used to train OpenAI models) and Microsoft Copilot with commercial data protection (enabled by default and recommended for faculty/staff, with chat history not saved and prompts/responses not used to train models). Personal/non-Edu ChatGPT accounts are not approved for sensitive or non-public data.