AI Tools, Teaching, and Academic Integrity at USI
Faculty members have had questions about whether the USI's current Academic Integrity policy applies to work that was created by artificial intelligence (AI) tools that generate content and strategies that could be used to mitigate cheating and enhance student learning experiences.
We have compiled information that we believe will be useful. This was developed through a collaboration between the Provost's Office, Dean of Students Office, CETL, and Online Learning, and was reviewed by the Provost’s Council in April 2023.
As AI technology and the higher education landscape continue to evolve, we will regularly review USI’s guidance and policies.
FAQs: Frequently Asked Questions on AI and Teaching
A1: Yes, yes, and yes. The Student Rights and Responsibilities: Code of Student Behavior describes academic integrity, including cheating, fabrication, and plagiarism. Unauthorized use of generative AI tools for submissions (such as essays, text, code, images) or improper citation or acknowledgment of such assistance would be violations of academic integrity since submissions would be misrepresented as “works as their own.”
The Dean of Students Office is reviewing the Student Rights and Responsibilities: Code of Student Behavior to identify any areas for clarification regarding the use of generative AI tools.
A2: Please see the suggested language in the Syllabus Statements (and copied below) for courses in which AI is forbidden or authorized with the instructor’s prior permission or direction.
The suggested syllabus language below comes from Duke University, the University of Iowa, Colorado State University, and Salem State University. See additional examples of syllabus statements and course policies in this crowd-sourced compilation of Classroom Policies for AI Generative Tools.
Suggested syllabus statements
We recommend that instructors include the first paragraph below, which provides an overview, and then decide on the extent of AI tools use in your course to add an appropriate description (2 examples are provided). [Do not copy the notes to instructors displayed within brackets.]
AI Tools Use in This Course (recommended)
Artificial intelligence (AI) tools that generate text, images, code, and other content are widely available. If you submit work containing any content generated by AI when not explicitly allowed and not in a way directed by me, the instructor, then this will be considered academic dishonesty and a violation of USI’s academic integrity policy. If you are not sure about what may be academic dishonesty or plagiarism and what is acceptable use in this course and on specific assignments, please contact me to discuss.
[Instructor: Determine the level of acceptable use of AI tools and include syllabus language to clarify this for your students. Suggested statements for two common scenarios are below.]
[Example 1, as applicable] AI use is prohibited.
You are not permitted to use AI tools that generate content (such as ChatGPT, Bing Chat, Bard, DALL-E) for work done for this class. Writing, analytical, and critical thinking skills are part of your learning outcomes in this course; therefore, all writing assignments should be prepared by you, the student. Content created by AI tools may not be considered your own original work.
This course assumes that work submitted by students (all process work, drafts, final versions, and all other submissions) will be generated by the students themselves, working individually or in groups (as directed).
[Example 2, as applicable] AI use only with prior permission or direction.
In general, the use of AI tools that generate content (such as ChatGPT, Bing Chat, Bard, DALL-E) is not permitted for work done for this class, except for specific assignments that I have identified and given specific guidelines for appropriate use of AI tools. All work submitted in this course must be your own.
Contributions from anyone or anything else (including generative AI tools such as ChatGPT, when permitted) must be properly quoted and cited every time they are used. When permitted for use in specific assignments, you must clearly identify the use of generative AI tools in your submission by citing which AI tool was used and the prompts you used to generate the content.
A3: Instructors are encouraged to engage in discussions with their students about USI’s academic integrity values and policies.
Students benefit from direct and transparent instructions about which tools and resources they can appropriately use in their work (Winkelmes et al., 2016). They will also need guidance on how to cite or acknowledge the generative AI tools, sources, and individuals who contributed to their work (as appropriate).
While a policy in a course syllabus is a good way to start these conversations, the syllabus shouldn’t be the only time that the policy and expectations are discussed with your students. Instructors are encouraged to have discussions throughout the semester about issues of AI tools that generate content, such as the benefits and shortcomings, ethical and privacy considerations, and the role of such tools in the profession.
A4: We suggest revisiting fundamental questions in your course and assignment design: What are the purpose and goals of the assignments and assessments? What do you want your students to practice and demonstrate? How do these goals align with the learning outcomes? Likely, you intend for your students to go beyond finding information. It might be helpful to focus on the writing process, which includes outlining, drafting, revising, and/or peer editing.
As you design your assignment, consider using specific prompts that ask students to incorporate class-specific or local content, recent information, or personal experiences, or compare/contrast and analyze different sources or styles of content. Your assignment could include an annotated bibliography and require citation of all their sources resources; if generative AI tools such as ChatGPT are allowed for a portion of the assignment, then show students how to appropriately cite or acknowledge their use of these tools (see examples in this crowd-sourced compilation of Classroom Policies for AI Generative Tools).
A5: There is not one fully reliable way. For example, as you get to know your students, you might become familiar with their style and tone. You also can check your writing prompt on ChatGPT and revise as needed, if you are comfortable with using such generative AI tools.
One consideration, expressed by Marc Watkins who is an instructor at the University of Mississippi, “We should be proactive in our response and not approach our teaching out of panic and mistrust of our students. What message would we send our students by using AI-powered detectors to curb their suspected use of an AI writing assistant, when future employers will likely want them to have a range of AI-related skills and competencies?” (Watkins, 2022).
AI detectors and their limitations:
Generative AI tools, such as ChatGPT, were only widely launched and available in the Fall of 2022. This is such a new tool that there has not been sufficient time to develop detection tools to detect AI-generated content. While there are some tools marketed for AI-generated text detection, they are too new to be accurate. For example, the U.S. Constitution was analyzed by GPTZero, an AI detector, and the tool determined that a large portion of the Constitution was likely written by AI.
SafeAssign (which is available in Blackboard) is used as a tool to determine if students are potentially submitting work as their own that has been previously submitted by themselves or is the work of others. SafeAssign utilizes a database of previously submitted assignments across institutions using this tool to determine how closely a submission matches something previously submitted.
SafeAssign does not have a tool to determine if work submitted has been generated by an AI-generation tool, such as ChatGPT. The creation of AI-generated work is unique and will not provide a match to an assignment in the database. Currently, there are no discussions about SafeAssign adding an AI detection tool, with an initial test by beta users during Summer 2023. There is no release date yet for all users. SafeAssign is partnering with ChatGPT to develop this detection tool.
- A Plagiarism Detector Will Try to Catch Students Who Cheat With ChatGPT (Chronicle, Surovell, 4.03.23)
- Can Turnitin Cure Higher Ed’s AI Fever? (Inside Higher Ed, Knox, 4.03.23)
This page was inspired by and based on AI and the CSU Student Conduct Code (Colorado State University) and Artificial Intelligence Tools and Teaching (University of Iowa).
Additional Resources on AI Tools and Teaching
- AI and Teaching - Guidance for Instructors (Duke University; 9 min read)
- How to Productively Address AI-Generated Text in Your Classroom (Indiana University, January 2023; 15 min read)
- Practical Responses to ChatGPT (Montclair State University, March 2023; 8 min read)
- Adapting Your Teaching to AI Generator Tools (Holton, 2023; 4 min read)
- Artificial Intelligence Writing (University of Central Florida, 2023; 4 min read)
- Handout for students: Independent Versus AI-Assisted Learning (Sharma, February 2023; 6 min read)
- Strategies for Teaching Well When Students Have Access to AI Text Composition Tools (George Mason University)
- Will ChatGPT get you caught? Rethinking of Plagiarism Detection (Khalil and Er, 2023)
- ChatGPT and Artificial Intelligence in higher education: Quick start guide (UNESCO, April 2023)
- Classroom Policies for AI Generative Tools (crowd-sourced)
- AI in Education Resource Directory (crowd-sourced)
- ChatGPT resources: A crowd-sourced collection of 276+ articles, blogs, podcasts, videos
- ChatGPT and Assessment at USI: Follow-up Conversations (USI CETL, March 2023)