Univeristy of Pennsylvania
Close

Academic Integrity Statements that Address Generative AI

25Jul’23

Syllabus Languages & Policies

Generative AI in Your Academic Integrity Statement

The evolution of generative AI brings considerations and concerns about how to ensure that students use these tools ethically in relation to their coursework. Generative AI can be powerful in improving student writing, research, brainstorming and organization.  In addition, students will likely be using these tools to generate text and visual content throughout their working lives. 

As a result, faculty have a responsibility to set clear guidelines for students around using generative AI in individual courses or assignments. Faculty may forbid students from using it or allow certain uses, such using it to revise a draft or generate topics, perhaps requiring students to disclose their use. No matter what you choose, clarity is crucial to helping students know what is considered academically honest use and what is not.

Sample Statements

Jonathan Zimmerman, HIST 0012, Why College? Historical and Contemporary Perspectives

If you use generative AI to do your thinking for this class, you'll be wasting a huge (and expensive!) opportunity to learn how to think and communicate. You can cut corners if you wish, but you will regret losing the opportunity to develop your own thoughts and skills. 

Holly Fernandez-Lynch, BIOE 5540: Bioethics and the Law

You may use AI programs, e.g., ChatGPT, to help generate ideas and brainstorm.  However, you should note that the material generated by these programs may be inaccurate, incomplete, or otherwise problematic.  Beware that use may also stifle your own independent thinking and creativity.

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 Penn policies.

Ryan Baker EDUC 6191: Core Methods in Educational Data Mining 

Within this class, you are welcome to use foundation models (ChatGPT, GPT, DALL-E, Stable Diffusion, Midjourney, GitHub Copilot, and anything after) in a totally unrestricted fashion, for any purpose, at no penalty. However, you should note that all large language models still have a tendency to make up incorrect facts and fake citations, code generation models have a tendency to produce inaccurate outputs, and image generation models can occasionally come up with highly offensive products. You will be responsible for any inaccurate, biased, offensive, or otherwise unethical content you submit, regardless of whether it originally comes from you or a foundation model. If you use a foundation model, its contribution must be acknowledged in the handin; you will be penalized for using a foundation model without acknowledgement. Having said all these disclaimers, the use of foundation models is encouraged, as it may make it possible for you to submit assignments with higher quality, in less time.

The university's policy on plagiarism still applies to any uncited or improperly cited use of work by other human beings, or submission of work by other human beings as your own.

Kate Dorsch STSC 1101: Science & Literature

As historians, sociologists, and philosophers of science and technology, we are always watching, studying, and thinking about the promise and pitfalls of new technologies - for ourselves, for our institutions, and for our societies.

I recognize that generative AIs are becoming a part of our lives. That includes as part of our work here at Penn and I recognize that you may use generative AI as part of your study practice in my class and many others.

It is important to remember that generative AI is a tool. Its efficacy and productive value depends directly on how it is used. It will not give you the best paper, the most fulsome explanation, the most reliable literature review. (Many new iterations of these generative AIs will completely fabricate information to complete the task given.) It cannot think for you. 

I encourage you to explore the use of generative AI - in your life in general and in this course in specific. As scholars in Science and Technology Studies, we may have an intellectual obligation to do so in a critically sustained way! Thus:

  • You are allowed (though not required) to use generative AI in this course. No assignments will require generative AI, as a reflection of our very serious ethical concerns about data sharing. No member of our group should feel required or obligated to use generative AI; the use or non-use of the tool will not affect the grading rubric (as long as the requirements below are met). No assignment will be designed such that students who use generative AI will have an “upper hand” or unfair advantage.

  • If you are using generative AI, please indicate as much in your journal, assignments, and so on.

  • Up to 200 words of any assignment can and should be devoted to how you applied the tool to your work, and your own assessment of how useful generative AI was in that application. Quote the AI as you would any other source; include your prompts with the output. Attaching screenshots to assignments is fine! 

Please feel free to talk to me about any aspect of this! I am very interested in how students are using generative AI - as an instructor I like to keep up with what my students are doing. But more so, as a scholar who studies knowledge production and trust, I am super interested in how this tool is being used!

Jeffrey Babin, EAS 5450, Engineering Entrepreneurship

We acknowledge the growing prevalence of AI in today's academic and professional environment and the potential it holds for facilitating various tasks. However, in this course, AI should not be used to complete quizzes or essays. These tasks are designed to develop and evaluate your own understanding, critical thinking, and ability to articulate your thoughts and arguments.

For other written assignments, the use of AI tools is permitted as a supplement to your own ideas and research. AI can aid in information collection and organization, but the final output should reflect your own understanding, analysis, and synthesis of the topic. In every instance where AI is used, you are required to disclose its use and appropriately cite all sources, including AI-generated content or suggestions. Failure to do so will be considered a violation of the academic integrity policy of this course. The use of AI should augment your learning experience, not replace your own intellectual work. Your engagement with the course content is crucial in fostering a rich learning environment and developing critical skills in technology innovation, marketing, and product management.

NOTE: This AI policy was drafted using ChatGPT 4.0 from OpenAI. I input the course syllabus and asked ChatGPT to draft an AI policy for the course that allows use for written assignments, excluding quizzes and essays.

Travis McGaha, CIS 2400, Introduction to Computer Systems

This isn’t as much a policy as it is advice. For this class, I recommend against using generative AI like ChatGPT or related technologies. I am not denying that ChatGPT can be a useful tool, I am just not convinced it is useful for your growth and learning of the material in this class.

  • ChatGPT is pretty good at making things sound correct, in ways that can make it hard to discern when it is actually incorrect vs when it is spouting nonsense. (See ChatGPT is a bullshit generator. But it can still be amazingly useful). When you are first learning about a topic, it can be especially hard to determine what is correct and what is “incorrect but articulated in a way that sounds like it is correct”.

  • There are also issues with how correct GPT is in general. People with more expertise than I on this matter hypothesize that many claims of ChatGPT ability to pass exams and solve coding problems come from the fact that it memorized the solutions enough during training. In other words, ChatGPT is being evaluated on its training data, biasing the results. Tests on questions outside of its training data give a much more negative evaluation of ChatGPT’s abilities to solve problems. As a person, you are likely to encounter many issues that can’t be solved only via internet memorization (like some of the exam questions for this class). See GPT-4 and professional benchmarks: the wrong answer to the wrong question.

I do not claim that ChatGPT is useless by these statements. There are many things it can do, I just doubt their usefulness in this class. We will not ban the usage of ChatGPT or similar tools except for during exams, but we expect you to be able to:

  • Explain and reason about your code yourself during office hours, project milestones, and demos
  • Reason and think critically about exam questions without help. Questions will be designed for you to reason about them, not just to have memorized the correct answer.

All in all, use your best judgement.

POLICY ON GENERATIVE AI (i.e., Chat GPT): Generative AI is a useful and time-saving tool for certain specific tasks. It can and should be used for such tasks. As a tool, AI should not be viewed as a way to avoid learning and the course material, but rather as a assistant to better master content.

In this course, you may use generative AI programs as preparatory and proofing work: generating topics, brainstorming, proofreading. But be aware that the use of these tools also may well stifle your own independent thinking and creativity, and constrain engagement with the specific themes developed in our lectures. 

Please note: the material generated by these programs may be inaccurate, incomplete, or otherwise problematic. These tools have a tendency to ‘hallucinate’ references (present seemingly plausible accounts as real when they are entirely fictitious).  As a general rule, distrust any fact produced by a genAI source until it is cross-referenced with a reliable source. These tools poorly model specific course readings in this class and particular threads/themes we develop in lectures and recitations.

If you do use generative AI programs in preparatory work on an assignment, you must indicate in a note preceding the assignment’s bibliography where and how you used such tools. 

A core goal of this course is to develop critical and analytic thinking. Developing such skills is the purpose of a university education in general and training in History specifically. Writing is integral to thinking. It is also hard. Generative AI does not replace the development of these skills. I am here to teach you and to develop these skills.  

Written work created by generative AI tools is not original work. You may not submit any written work generated by an AI program as your own. All assignments should be your own original work, created for this class.