Skip to main content

Grand Challenge InitiativeArtificial Intelligence

Syllabus AI Policies

Policy Ideas for the Use of AI in Higher Education

Artificial Intelligence (AI) is one of the key transformative technologies of the coming century.  Its impact on research institutions and universities is already significant today, and is likely to continue growing intensively over the coming years.

For students and scholars, AI can be a powerful tool that enhances their learning experience or research endeavors – but, if used improperly, it can also degrade the quality of a student’s education and undermine the rigor and impact of scholarly research.  Much depends on the specific pedagogical or intellectual context within which AI tools are being used, and on the ways in which those tools either support or weaken the underlying goals of a given academic endeavor.  Since these contexts vary widely, it is the policy of Vanderbilt University to leave it up to individual faculty to decide whether and how AI tools are used in their courses or research work.

Below we include the most recent Guidance from Faculty Affairs, which stresses the importance of each faculty member deciding whether or how to incorporate generative AI into course expectations.

In an effort to assist faculty in making these their own policies, the AIGC has catalogued example policies regarding the academic uses of AI as falling along a broad spectrum that ranges from permissive to restrictive.  Here we present examples of such policies that have been adopted by Vanderbilt faculty hailing from a variety of disciplines.  We have divided these examples into four basic gradations: permissive; slightly restrictive; moderately restrictive; and completely restricted.

If you would like to submit your own policy to be included, please click here.

University Policy

Vanderbilt University Academic Affairs Guidance for Artificial Intelligence (AI) Tools

POLICY STATEMENT:

Artificial Intelligence is a valuable, dynamic, and vital tool. As with all other tools, you must abide by all applicable laws and policies when you use AI, and you should strive to use AI ethically. You should follow guidelines (including prohibitions) provided by your dean, instructor, supervisor, or other individual(s) overseeing your work. You should disclose the use of AI in an appropriate way. You should exercise sound judgment in your use of AI. When AI is used, it should be used in ways that are consistent with university policies on confidentiality and privacy. This applies to all aspects of research, service work, and creative expression. You are the author of content that you produce with AI and responsible for its accuracy, impact, and compliance with relevant laws and policies.

PRINCIPLES/GOALS:

  • Faculty should decide whether and how generative AI is used in courses.
  • Faculty should clearly communicate expectations to students.
  • Faculty should clearly communicate what constitutes academic dishonesty.
  • Students are responsible for understanding the rules of engagement for using AI in each of their courses and seeking out information if they do not understand or are unsure how to comply.
  • Faculty and students are responsible for using AI appropriately and ethically.
  • Faculty and students should disclose the use of AI in their work if such disclosure is expected.
  • Faculty and students are the authors of content generated by AI and are responsible for that content as they are with content that they author.
  • AI should be used in ways that respect confidentiality and privacy.
  • AI should be used legally, ethically, and reasonably.

Permissive Policies

Scott Crossley, Peabody

Generative AI. You can use generative AI models (ChatGPT, GPT, DALL-E, Stable Diffusion, Midjourney, GitHub Copilot, and anything after) as needed, for any purpose, at no penalty as long as you recognize the model’s contribution. Failure to acknowledge the use of AI models will be penalized as plagiarism. However, please note that AI models tend to include incorrect information, fake citations, and inaccurate outputs. You are responsible for any inaccurate, biased, offensive, or otherwise unethical content submitted regardless of whether it originally comes from you or an AI model.

Somewhat limited Policies

David Schlundt, Psychology

The worlds of science and academics are rapidly changing with the introduction of complex-language AI software. This software, like ChatGPT, can be very useful when researching topics. You are encouraged to explore AI tools to find and organize knowledge.  I strongly recommend that you avoid pasting AI output into your papers. You will have to backup what you write with citations to studies, literature reviews, and other scholarly works. Helping you become a better writer is one of the most important course goals.  You may use AI but write the papers using your own words and only use citations that you have accessed and reviewed for accuracy. All papers will be reviewed for plagiarism using Turnitin software. The software now includes the ability to detect text that was generated using AI.  Plagiarism is a violation of the Honor code. Currently, I will treat copying text from AI as a reason for deducting points to further encourage you to write in your own words.

Michael Bess, History

For the purposes of this course, the use of AI tools such as GPT4, Bing, Claude, or Bard falls under two categories:
1. text-generation tool (prohibited);
2. research, brainstorming, and editorial aid (permitted).
Turning to AI text-generation tools can all too easily degrade the quality of a
student’s learning experience in thinking through the material for themselves and creating their own work. On the other hand, certain targeted usages of AI tools can powerfully augment the learning experience for students, when conducted in a way that enhances their critical thinking, creativity, and research capabilities. Please read the detailed description of this policy (and the rationale for it) below. Failure to adhere to this policy, or to acknowledge your use of such a tool, constitutes a violation of Vanderbilt’s honor code. Feel free to consult me if you have any questions about this.

Prohibited use:

Unless I have explicitly instructed you to do so, I do not allow students to use AI text-generation tools in drafting the text of their essays, Reading Reports, Film Reports, or other assignments for my courses. This prohibition applies to all the following cases:

1. Entire AI-generated essay or assignment. A student instructs an AI text-generation tool to write an entire essay or assignment, then hands in the assignment as if it had been written by the student.

2. Partial AI-generated essay or assignment. A student instructs an AI text-generation tool to write a portion of an essay or
assignment, then hands in the assignment as if the entire assignment had been written by the student.

3. Modified AI-generated essay or assignment. A student extensively modifies an AI-generated essay in ways that result in a hybrid of the student’s own phrasing intermingled with AI-generated text, then hands in the assignment as if the entire assignment had been written by the student.

4. Paraphrased AI-generated essay or assignment. A student completely paraphrases an AI-generated essay or assignment in ways that result in a new text that is entirely written by the student, but that is merely a thorough rewording of a text generated by the AI. Such a paraphrased text still follows the same overall structure of organization as the AI-generated text, and closely echoes the principal ideas presented in the AI-generated text. The student then hands in the assignment as if the entire assignment had been researched, conceived, and organized by the student.

Permitted Use:

I do allow students to use AI tools in conducting background research for their assignments, for thinking through the ideas they wish to explore, and for editing the style, wording, and language of their assignments before handing them in. This allowable usage would include, for example, the following cases:

1. AI tools used for background research. A student consults an AI tool by asking it basic factual or thematic questions about a topic, then seeing what kinds of material the AI presents in response. This is similar to consulting Wikipedia at the outset of a project, in order to get a quick sense of the main factual and thematic contours of the subject matter.

2. AI tools used for brainstorming. A student consults an AI tool with questions about basic concepts, ideas, principles, theories, or scholarly debates relating to a topic the student wishes to explore. This is similar to consulting scholarly articles online, in order to get a sense of the main conceptual or theoretical contours of the subject matter.

3. AI-generated essay or text is consulted before student writes their own essay. A student prompts an AI text-generation tool to write an essay on a topic assigned for this course, but only uses the AI-generated text for “consultative” purposes, in order to see what kinds of ideas or arguments the AI has come up with. After reading the AI- generated essay, and reflecting critically about it, the student then conducts their own research and reflection about the topic, using the kinds of scholarly tools available to humans before the advent of advanced AI (for example, books, journal articles, online sources, debates with classmates, conversations with professors). The AI-generated essay thus becomes merely one element within the broad array of other resources that the
student consults, and critically reflects upon, in researching, debating, and crafting their own final product. (Note: with this usage, students need to be extra careful not to violate principle #4 above, namely, merely paraphrasing portions of the AI-generated text or closely echoing its organizational structure. The final product needs to be the student’s own work of critical reflection and synthesis.)

4. AI-edited version of student’s essay or assignment. A student composes their own entirely original essay or assignment, then submits the assignment to an AI tool to see what kinds of stylistic edits and/or grammatical modifications the AI recommends. The student then hands in the assignment after making some or all of the editorial changes recommended by the AI. Important: in all such permitted usages – that is, whenever you have used an AI tool in any way as you worked on an assignment – you must include a citation of this tool in your references, indicating in detail the ways in which you used this tool in researching and/or crafting your assignment. Failure to acknowledge your use of such a tool constitutes a violation of Vanderbilt’s honor code.

* * *

My rationale for this policy is based on a fundamental distinction: transactional assignments (shallow and weak) vs. transformative assignments (deep and enduringly impactful). A transactional assignment is one that is seen by the professor and/or student as a mere means to the end of getting a grade. It’s like a business transaction: the student assembles some information and submits it, and the professor assesses the quality of the submission and gives it a grade. Eventually, in this impoverished and sterile model of education, the student graduates from Vanderbilt with a certain GPA, which allows the student to go out in the world and get a job. Vanderbilt is little more than a hoop through which the student must jump in order to become more employable. The person who came here as a first-year student is not deeply affected by the courses they take, and does not end up being all that different from the senior who graduates four years later: all that they have done is to add an external certification to their credentials. A transformational assignment, by contrast, aims to launch both the student and professor on a journey of mutual personal growth and intellectual exchange and engagement. As the student works on the assignment, they struggle to articulate new ideas, to ask new kinds of questions, to find new critical perspectives from which to interpret the material at hand. This is hard work, because it necessarily entails a face-to-face encounter with ideas and questions that are strange and unfamiliar, taking the student through phases of bafflement and perplexity as well as the wonderful “aha!” moments that signal insight and innovation. The student emerges from this process as a partially transformed person, endowed with new critical skills and a modified vision of the world and their own place within it. Ideally, the professor will partake in this learning process as well, gleaning new perspectives from interacting with the student as the challenging creative process unfolds. In this vision of education, Vanderbilt becomes the site of a gradual, cumulative process of pivotal change in the student’s life – a place where not just new knowledge, but a newly enriched personal character, has come into being. Both the student and the professor have changed in meaningful and important ways, becoming noticeably different persons than they were before.

Not all uses of AI text-generation tools are superficial or transactional in their impact. In some classes, for example, I have done a flipped version of an essay assignment that deliberately incorporates an element of AI-generated text. I instruct students to go to an AI website such as ChatGPT, and to insert a prompt that causes the AI to write an essay on a given topic related to the course. The student’s assignment at that point is to write a critique of the essay generated by the AI, detailing both the strengths and the weaknesses of the AI-generated essay. This could mean, for example, reflecting on the ways in which the AI has pointed out new information or analytical points that may not have occurred to the student, but also other ways in which the AI may have failed to capture important nuances in the topic. This latter “flipped” use of the AI stays true to the pedagogical mission of our university, for two reasons: first, the role played by the AI is openly acknowledged. Second, the assignment is designed to enhance the opportunities for the student’s critical reflection on the subject matter: the AI-generated essay becomes just one more element in the student’s learning process as they grapple with the substantive issues underlying the essay assignment as a whole. The student is being asked, in effect, to conduct a “peer review” of the AI-generated essay, and this exercise can be a powerful learning tool. The tools of AI, in this sense, are not all that different from other learning tools such as books, lectures, podcasts, or conversations. They can be used in a shallow and transactional manner, as shortcuts toward “getting an assignment done.” Or they can be used as vehicles for transformative practice, critically marshalled in a more difficult and disruptive form of learning that shakes up your world and leads you into new personal and intellectual territory. 

Unattributed (adapted from Monica Linden, Brown University).

The use of AI has exploded in just this last year. Although many of the assignments in this class will not lend themselves to use AI (I think) because a lot of it is personal reflection, it’s possible that it could be used in some instances. So for our purposes, I ask that you report your use, using the following guidelines. There will be no deduction in your grade for the reported use of AI as this is an experimental time. At the bottom of your assignments please write the statement that most closely describes your use. If you have used it extensively, I will call you in for a conversation on a case-by-case basis.

  • I did not use Generative AI for this assignment.
  • Generative AI helped with grammar and readability, but the original writing was my own.
  • Generative AI helped with idea generation, but I wrote the assignment myself.
  • A small amount of my work (just a couple of sentences) was written by Generative AI.
  • Generative AI significantly wrote my work, but I edited it.

An example of unacceptable use of AI would be to have it write a song on a particular topic without making significant changes. If you do use it to give you some ideas, you will need to submit a screenshot of the prompt and the song the AI wrote along with your assignment.
If you do use AI in your work, you must properly cite it. Look at these guidelines for how to cite properly in APA style: https://apastyle.apa.org/blog/how-to-cite-chatgpt

I reserve the right to modify the guidelines as needed.

Ole Molvig, History/CSET

In this course, students are allowed limited and intentional use of AI text-generation tools such as ChatGPT, Bing, or Bard in crafting essays or other assignments for this course.  (The only exception is for specific assignments in which we explicitly instruct students to use such a tool.)

Care should be taken in the use of AI for education.  Turning to AI text-generation tools can all too easily degrade the quality of your own learning experience in thinking through the material for yourself and writing your own essay.  Therefore it is not allowed to simply use AI text-generation tools such as ChatGPT, Bing, or Bard merely to replace your own effort.  (In other words, you may not simply enter the essay prompt, hit generate, and turn the work in as your own.)  AI tools may be employed in the brainstorming, research, and editing of class assignments, but the final, submitted essay must be primarily written by the student author.  Text predominantly generated by an AI must be cited as such, as you would a human author.  Note, rarely will an AI expert be a preferred citation to a human expert.

In the use an AI text-generation tool, you must include a citation of this tool in your references, indicating in detail the ways in which you used this tool in crafting your assignment.  Failure to acknowledge your use of such a tool constitutes a violation of Vanderbilt’s honor code.  For a full description of this policy (and the rationale for it), please see the statement in the class pack; do feel free to reach out to us if you have any questions about this policy.

Richard Alan Peters II, Electrical and Computer Engineering

Use of generative AI.
You are free to use generative AI algorithms such as Chat-GPT in your work. However, You must:

1. Cite any text that the AI generated (even if you edited it) with a bibliography entry that includes the name and version of the AI model that you used, the date and time it was used, and includes the exact query or prompt that you used to get the results.
2. Cite, as described in rule 1, any code that you had it generate for you.I recommend that you not ask it to write code for you. Doing so will probably be more work than simply writing it yourself. Because:
(a) You must thoroughly test the code to prove that it works.
(b) You must explain what you did to verify that it works.
(c) To demonstrate that you understand it, you must comment every single logical object be it a data structure or line or short block of code that it generates. I.e. exactly what that bit of code does and how it does it.
(d) The code must follow the other rules. For example the assignment may have stated restrictions on methods, procedures, external libraries, or programs.
(e) It must generate the results that are asked for in the assignment instructions.

I hope that by following these rules you will learn how to use generative AI as
an assistant to increase your productivity in writing and coding. If you fail to
follow these rules, that will be an honor code violation and you will be referred
to the Honor Council.

Strongly limited Policies

Unattributed

You may use AI programs e.g. ChatGPT to help generate ideas and brainstorm. However, you should know that the material generated by these programs may be inaccurate, incomplete, or otherwise problematic. You may not submit any work generated by an AI program as your own. If you include material generated by an AI program, it should be cited like any other reference material (with due consideration for the quality of the reference, which may be poor). Any plagiarism or other form of cheating will be dealt with severely under relevant Vanderbilt policies. Of course, there are many cases where I may not be able to tell if you cheated or not. Note however that, for any reason, including doubts about whether the paper was written with external help, any student can be required to come in for an oral examination on their paper. In such a case, a grade for the oral examination will replace that for the paper.

Anna Castillo, Spanish

Policy regarding ChatGPT and other AI tools: The objective of this course is to help you to develop your own voice as a film critic while cultivating your intellectual creativity. The use of generative AI to substantially complete your homework goes against this objective and the honor code. Having said this, learning to use artificial intelligence (AI) language models is an emerging skill, and I can give you advice about using it in a responsible and effective manner. If you have any doubts, ask me. It is important not to outsource your learning in this course. (translated from Spanish)

Ray Friedman, Owen

You are writing about your own personal experiences, which are not in the AI databases.  Therefore, chat bots cannot answer the assignments.  Chatbots can sometimes help with writing.  If you would like to use AI to help your writing, please submit a) your own version (an outline, at least) and b) the AI-enhanced final version.  I will assess your final version only but want to also have a clear sense of your initial input.  (Careful! AI can also sometimes produce vapid buzz-word-filled sentences that name no sense.  It is up to you to decide what words best represent your ideas.)

Restricted Policies

Unattributed

Due to the centrality of the craft of writing in this course, students may use NO AI text-generation tools (such as ChatGPT, Bing, or Bard, or similar) in any way in crafting their essays or other assignments for this course.  Turning to AI text-generation tools can all too easily degrade the quality of the student’s learning experience in thinking through the material for themselves and creating their own essay.  The use of such a tool constitutes a violation of Vanderbilt’s honor code.  Feel free to reach out to me if you have any questions about this policy.