Teaching & Learning
- Additionally, for undergraduates, using non-Princeton tutoring resources (including AI tutoring bots) when completing any work to fulfill an academic requirement is prohibited
- Princeton states that generative AI use for academic work is governed by course and departmental policies, and students are responsible for determining what is allowed in a given course and staying within instructor-specified parameters
- The University defines certain AI-related behaviors on work submitted to fulfill an academic requirement as academic integrity violations, including copying AI output, representing AI-derived output as one’s own, exceeding the instructor’s parameters, or failing to disclose AI use
Students are responsible for familiarizing themselves with and adhering to course and departmental policies regarding the use of generative AI.
Inappropriate uses of the results of generative AI on any work submitted to fulfill an academic requirement, such as directly copying the output, representing output generated by or derived from generative AI as their own, exceeding the parameters specified by the instructor, or failing to disclose its use, would constitute violations of academic integrity.
[U]ndergraduate students may use only Princeton-authorized academic support. When completing any work to fulfill an academic requirement (such as homework assignments, problem sets, quizzes or examinations of any kind), undergraduate students may not make use of any non-Princeton tutoring resources, including paid or unpaid tutors, independent tutoring services, or AI tutoring bots.
- Princeton’s Academic Regulations prohibit undergraduate students from using AI tutoring bots when completing any work to fulfill an academic requirement, and explicitly includes examinations as examples of such work
- More generally, the University states that inappropriate uses of generative AI results on any work submitted to fulfill an academic requirement (which may include assessments) can constitute academic integrity violations, including failing to disclose or exceeding instructor parameters; however, exam-specific AI rules beyond this are not uniformly defined and may depend on course policies
When completing any work to fulfill an academic requirement (such as homework assignments, problem sets, quizzes or examinations of any kind), undergraduate students may not make use of any non-Princeton tutoring resources, including paid or unpaid tutors, independent tutoring services, or AI tutoring bots.
Inappropriate uses of the results of generative AI on any work submitted to fulfill an academic requirement, such as directly copying the output, representing output generated by or derived from generative AI as their own, exceeding the parameters specified by the instructor, or failing to disclose its use, would constitute violations of academic integrity.
- For undergraduates, Princeton prohibits using non-Princeton tutoring resources (including AI tutoring bots) when completing any work to fulfill an academic requirement
- Princeton also emphasizes that students must follow course and departmental policies on generative AI and that disclosure is required when generative AI is permitted by the instructor
- Separately, Princeton provides institutionally supported access to Microsoft Copilot and recommends using it with a Princeton account for better data protection, while stating that the consumer version is not recommended for Princeton work or learning
[U]ndergraduate students may use only Princeton-authorized academic support. When completing any work to fulfill an academic requirement (such as homework assignments, problem sets, quizzes or examinations of any kind), undergraduate students may not make use of any non-Princeton tutoring resources, including paid or unpaid tutors, independent tutoring services, or AI tutoring bots.
If generative AI is permitted by the instructor (for brainstorming, outlining, etc.), students must disclose its use rather than cite or acknowledge the use, since it is an algorithm rather than a source.
Students are responsible for familiarizing themselves with and adhering to course and departmental policies regarding the use of generative AI.
Microsoft Copilot, a generative artificial intelligence (AI) tool that can be used to generate or analyze text and images, is now available for use at Princeton University.
The “Copilot for consumers” version is not recommended for Princeton work or learning.
- Princeton also provides course-policy examples indicating that in at least some courses AI may be used for debugging code only with required disclosure, and that using AI to generate code or complete problems may be prohibited by the instructor
- Princeton’s University-wide Academic Regulations treat inappropriate use of generative AI on any work submitted to fulfill an academic requirement (including representing AI-derived output as one’s own or failing to disclose) as an academic integrity violation, with specific allowances depending on course and departmental policies
Students are responsible for adhering to course and departmental policies regarding the use of generative AI. Inappropriate uses of generative AI on any work submitted to fulfill an academic requirement, such as directly copying the output, representing output generated by or derived from generative AI as their own (including rewording or paraphrasing output from generative AI and representing it as their own), exceeding the parameters of use specified by the instructor, or failing to disclose its use (including the use of generative AI for rewording, paraphrasing or editing) constitute violations of academic integrity.
You may use AI on problem sets to check for errors and debug code, and – if you do – you must disclose this use, naming the tool (e.g., “checked for errors using Copilot”). You may not make use of AI to generate code or complete the problems.
Research
- Beyond these statements, research-manuscript-specific rules are not defined in the provided text
- Princeton’s Microsoft Copilot guidance recommends avoiding prompts that include restricted information, including proprietary research data, and instructs users to review and validate AI outputs before relying on or sharing them
- Princeton’s OIT indicates that the Generative AI in Research Advisory Group produced guidelines for the use of generative AI in academic research and related activities, but the OIT page itself does not state specific required/prohibited practices for research writing
The Generative AI in Research Advisory Group explored the need for the development of resources to support scholars aiming to enhance research practices with generative AI. The group has generated AI resources for researchers and guidelines for the use of generative AI in academic research and related activities.
It is recommended that you avoid writing prompts that include restricted information, including proprietary research data.
Regardless of the tool or mode, AI responses can be biased, inaccurate, inappropriate, or may contain unauthorized copyrighted information. Always review, validate, and iterate on responses provided by any Generative AI tool before relying on them or sharing them broadly.
- Specific rules about AI use for research data analysis beyond these tool and data-handling statements are not defined in the provided text
- Princeton’s OIT lists Princeton-supported AI tools and states that Princeton provides an AI Sandbox for Faculty Research in a secure environment for faculty researchers to use multiple LLMs
- For data protection, Princeton’s Copilot guidance recommends avoiding prompts that include restricted information, including proprietary research data, and warns that using Copilot without a Princeton account or using another tool not provided by OIT can result in submitted information being used to train the model and potentially becoming available outside Princeton
Princeton provides a secure environment for faculty researchers to make use of multiple Large Language Models (LLMs).
It is recommended that you avoid writing prompts that include restricted information, including proprietary research data.
If you plan to use Copilot without your Princeton.edu account or another Generative AI tool that has not been provided by OIT, please ensure that the information you submit is unrelated to Princeton work or learning.
This submitted information can be used to train the underlying model and potentially become available to users outside of Princeton University.
- In the OIT generative AI guidance, Princeton also emphasizes transparency and the need to review and validate AI-generated content
- The OIT “For Research” page states that a Generative AI in Research Advisory Group generated guidelines for the use of generative AI in academic research and related activities, but does not provide the guideline text itself
- Princeton’s Graduate School regulations state that academic and research integrity expectations extend beyond formal degree requirements to professional activities undertaken by graduate students, and that failure to abide by responsible conduct of research standards can be treated as a serious violation subject to disciplinary action
For graduate students, scholarly and research activity that is subject to the University’s standards may go beyond the work submitted in fulfillment of an official academic requirement for the degree.
Academic and research integrity is an expectation that extends to all professional activities undertaken by graduate students on or off campus.
Additionally, failure to abide by standards related to the responsible conduct of research when students ought reasonably to understand those standards will also be considered a serious violation and will be subject to disciplinary action.
The Generative AI in Research Advisory Group explored the need for the development of resources to support scholars aiming to enhance research practices with generative AI. The group has generated AI resources for researchers and guidelines for the use of generative AI in academic research and related activities.
Transparency: Clearly communicate to the end users and stakeholders the use of AI to foster an environment of trust and accountability.
Accuracy and Quality: Maintain oversight and always review and validate AI-generated content to ensure accuracy and appropriateness.
Academic Integrity
- Princeton states that failing to disclose generative AI use (including for rewording/paraphrasing/editing) constitutes an academic integrity violation
- Princeton requires disclosure (not citation/acknowledgment) of generative AI use when the instructor permits it, because it is defined as not being a source
- Princeton also provides a disclosure template for students to use and emphasizes that students must determine whether, how, and to what extent a course allows generative AI before using it
As defined in section 2.4.7, generative artificial intelligence (AI) is not a source, since its output is not produced by a person. If generative AI is permitted by the instructor (for brainstorming, outlining, etc.), students must disclose its use rather than cite or acknowledge the use, since it is an algorithm rather than a source.
Students are responsible for adhering to course and departmental policies regarding the use of generative AI. Inappropriate uses of generative AI on any work submitted to fulfill an academic requirement, such as directly copying the output, representing output generated by or derived from generative AI as their own (including rewording or paraphrasing output from generative AI and representing it as their own), exceeding the parameters of use specified by the instructor, or failing to disclose its use (including the use of generative AI for rewording, paraphrasing or editing) constitute violations of academic integrity.
When generative AI is permitted, your second responsibility is to always explicitly disclose your use of it.
A Template for Disclosing AI Usage
"This document was created with assistance from AI tools. Specifically, I used [insert which AI tool] for [insert purpose: brainstorming, outlining, generating content for sections of text on pages (insert pages), revising or proofreading text, formatting citations, etc.]. I have reviewed and edited this work and have retained records of my generative AI use. For more information on the extent and nature of AI usage, please contact me.”
- The Information Security Policy states that failure to comply may lead to disciplinary action
- For detection, Princeton guidance for instructors states that it does not recommend using AI detection software (e.g., Turnitin, ZeroGPT, OpenAI detection capabilities) to determine whether student work is AI-generated
- Princeton’s academic regulations state that AI-related misconduct (including representing AI output as one’s own or failing to disclose AI use) constitutes violations of academic integrity, and that violating tutoring/online solutions provisions subjects students to disciplinary action
Inappropriate uses of generative AI on any work submitted to fulfill an academic requirement, such as directly copying the output, representing output generated by or derived from generative AI as their own (including rewording or paraphrasing output from generative AI and representing it as their own), exceeding the parameters of use specified by the instructor, or failing to disclose its use (including the use of generative AI for rewording, paraphrasing or editing) constitute violations of academic integrity.
Any student violating these provisions is subject to disciplinary action.
Failure to comply with this policy may subject you to disciplinary action and to potential penalties described in Section 1.1.7 of Rights, Rules, Responsibilities.
Though companies like Turnitin, ZeroGPT, and OpenAI have all developed AI detection capabilities, we do not recommend you use such software to attempt to determine if student work is AI-generated.
Institutional & Administrative
- Princeton’s general OIT guidance also requires oversight: users are instructed to review and validate AI-generated content and to protect sensitive information
- For teaching, Princeton states it does not intend to ban ChatGPT or impose a top-down edict on how each instructor addresses AI in classes, and encourages instructors to be explicit about their course policies
- Princeton’s Generative AI Tools Use Policy applies to individuals granted access to University Information (including employees, contractors, and others) and directs that only University-licensed generative AI tools should be used with University Information classified as Unrestricted Within Princeton or Confidential; using publicly available generative AI tools with such information is not permitted
This policy applies to all individuals and entities granted access to University Information, including, but not limited to, employees, contractors, temporary employees, and volunteers.
University-licensed generative AI tools should be used with Princeton Information classified as Unrestricted Within Princeton or Confidential under the University’s Information Security Policy. The use of publicly available generative AI tools in conjunction with such Princeton Information is not permitted by the University.
Accuracy and Quality: Maintain oversight and always review and validate AI-generated content to ensure accuracy and appropriateness.
Data Privacy: Protect sensitive information and adhere to relevant data protection regulations, and university policies, including data classification guidelines when crafting input prompts or giving Gen AI tools the access to documents, texts or images to an AI system.
Princeton University doesn’t intend to ban ChatGPT or to levy a top-down edict about how each instructor should address the AI program in your classes.
Be explicit about your AI/ChatGPT policy in your syllabus, on Canvas, and as you introduce assignments during your first (and perhaps subsequent) class meetings.
- Princeton further cautions that if using Copilot without a Princeton account or using another AI tool not provided by OIT, the submitted information can be used to train the model and potentially become available outside Princeton
- Princeton’s Generative AI Tools Use Policy prohibits using publicly available generative AI tools with University Information classified as Unrestricted Within Princeton or Confidential, and states that only University-licensed generative AI tools should be used with such information
- Princeton’s OIT generative AI guidance also states: “Don't use non-public Princeton data in public Gen AI tools.” Princeton identifies Princeton-supported AI tools (including Microsoft 365 Copilot Chat and an AI Sandbox for Faculty Research), and notes that Copilot with Enterprise Data Protection (when logged in with a Princeton account) provides protections including that prompts/responses are not used to train underlying models and chat data is encrypted; the consumer version is not recommended
Only generative AI tools licensed by the University should be used with University Information classified as Unrestricted Within Princeton or Confidential information under the University’s Information Security Policy.
The use of publicly available generative AI tools in conjunction with such Princeton Information is not permitted by the University.
Data protection: Don't use non-public Princeton data in public Gen AI tools.
Copilot is currently the only generative AI tool made available by the Office of Information Technology (OIT).
In order to help secure Princeton data and reduce risk to the University, log in to Copilot with your Princeton University account, to enable the following protections:
(a) your prompts and responses are not used to train the underlying large language models, and
(b) chat data sent to and from Copilot is encrypted.
If you plan to use Copilot without your Princeton.edu account or another Generative AI tool that has not been provided by OIT, please ensure that the information you submit is unrelated to Princeton work or learning.
This submitted information can be used to train the underlying model and potentially become available to users outside of Princeton University.
A generative artificial intelligence (AI) tool that can be used to generate or analyze text and images.
Princeton provides a secure environment for faculty researchers to make use of multiple Large Language Models (LLMs).
- Princeton’s OIT generative AI site states it will be regularly updated as the field evolves
- Beyond these statements, a detailed university-wide AI strategy/roadmap is not defined in the provided source text
- Princeton’s OIT indicates that the Generative AI in Research Advisory Group explored resource needs and generated AI resources and guidelines for researchers, reflecting an organized governance/support effort for research-related AI use
The Generative AI in Research Advisory Group explored the need for the development of resources to support scholars aiming to enhance research practices with generative AI. The group has generated AI resources for researchers and guidelines for the use of generative AI in academic research and related activities.
Note: As the field of generative AI continues to evolve, this site will be regularly updated.
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 Princeton University's AI Policies
Princeton University has defined AI policies in 12 of 12 categories, with an overall coverage score of 100%.
Princeton requires disclosure (not citation/acknowledgment) of generative AI use when the instructor permits it, because it is defined as not being a source. Princeton states that failing to disclose generative AI use (including for rewording/paraphrasing/editing) constitutes an academic integrity violation. Princeton also provides a disclosure template for students to use and emphasizes that students must determine whether, how, and to what extent a course allows generative AI before using it.
Princeton’s academic regulations state that AI-related misconduct (including representing AI output as one’s own or failing to disclose AI use) constitutes violations of academic integrity, and that violating tutoring/online solutions provisions subjects students to disciplinary action. The Information Security Policy states that failure to comply may lead to disciplinary action. For detection, Princeton guidance for instructors states that it does not recommend using AI detection software (e.g., Turnitin, ZeroGPT, OpenAI detection capabilities) to determine whether student work is AI-generated.
Princeton’s Generative AI Tools Use Policy prohibits using publicly available generative AI tools with University Information classified as Unrestricted Within Princeton or Confidential, and states that only University-licensed generative AI tools should be used with such information. Princeton’s OIT generative AI guidance also states: “Don't use non-public Princeton data in public Gen AI tools.” Princeton identifies Princeton-supported AI tools (including Microsoft 365 Copilot Chat and an AI Sandbox for Faculty Research), and notes that Copilot with Enterprise Data Protection (when logged in with a Princeton account) provides protections including that prompts/responses are not used to train underlying models and chat data is encrypted; the consumer version is not recommended. Princeton further cautions that if using Copilot without a Princeton account or using another AI tool not provided by OIT, the submitted information can be used to train the model and potentially become available outside Princeton.