Address the Impact of AI

Update your syllabus to address the use of AI tools, and discuss your expectations with students.

Students are more likely to cheat if they don’t know or understand the rules and conventions around academic integrity.

We’re just starting to imagine the ways that AI text generation tools might play a role in students’ writing processes. Text generation tools might, for instance, be useful to help students generate ideas in the brainstorming stage of writing, or to create simplified summaries of their rough brainstorming notes.

In a discussion of ChatGPT facilitated by OTEAR, Rutgers faculty expressed a range of views about what uses of AI writing tools might be acceptable. This is not surprising: the rules and conventions around academic integrity—and authorship more broadly—are evolving as we explore what these tools can and cannot do.

In light of this reasonable disagreement, you should state specifically how and whether AI may be used in your class, both in your syllabus and in discussion with your students.

A few possibilities for syllabus language:

• Use of AI such as ChatGPT is not permitted in any stages of the writing process on any assignment.
• Use of AI such as ChatGPT is only permitted to help you brainstorm ideas and see examples. All material you submit must be your own.
• Use of AI such as ChatGPT is fully permitted, but you must cite the tool and be able to explain any work that you submit.

For more detailed suggestions, please see this resource (scroll down to section 4, "Suggestions for Updating Your Syllabus").

Lead a conversation to set norms and commitments early in the semester; include AI tools in this discussion.

Students are more likely to cheat if they think that their peers are cheating, or approve of cheating. (This is among the most well-established factors contributing to student cheating behaviors.)

A collective norm and commitment exercise can address this. Do this activity just after add/drop period ends, when you know that your class enrollment is settled. It involves two parts:

1. A facilitated conversation or activity focusing on class norms, policies, and expectations—including, but not limited to, academic integrity. In a large course, this might involve a presentation from the instructor followed by Q&A; in a smaller course, this might involve students providing feedback and suggestions on specific class policies or practices.
2. A mechanism for the agreed-upon norms to be recorded and students to provide consent or express their commitment.

Consult this resource from the Stanford Teaching Commons for detailed guidance on setting norms and commitments in your class.

Collectively setting norms and making commitments promotes transparency, equity, and classroom community. In the context of concerns about inappropriate use of AI writing tools or other resources, the sense of community and accountability generated by such an activity can help to cultivate a sense among students that their peers disapprove of, and will not use, unapproved forms of assistance on their coursework.

Know what to do if you think someone has used AI in an unauthorized way.

There’s been a lot of news about AI writing detectors, but the risk of false positives is significant. Some of these tools also appear to be biased against nonnative english speakers. We have also heard anecdotal reports that there are relatively simple ways to evade most of the existing detectors.

In general, we do not recommend using AI detectors, and we strongly recommend against submitting an academic integrity concern solely on the basis of the output of an AI detector.

We recommend approaching suspected academic dishonesty involving AI writing tools in the same way as you would approach a case of suspected contract cheating: review the submitted work in the context of the course and the student’s other work. Does the submitted work have a different writing style or show a different level of comprehension than other work the student has submitted? Does the submitted work reference facts or ideas that were not discussed in class? If you find such evidence, you may submit an academic integrity concern.

Be transparent in your assignments. Clearly explain the assignment purpose, as well as the task and criteria.

Students are more likely to cheat if they don’t see the value in doing the work.

Students sometimes have difficulty perceiving the connection between an assignment and their learning process or their goals. To make your assignments more transparent, focus on explaining an assignment’s value to students’ learning; how it connects to course learning goals; and how the skills and knowledge that students develop will help them in their academic, personal, or professional lives.

Look at the “Example assignments” section of this page for sample transparent assignments from STEM, social science, and humanities disciplines.

You might also consider sharing with learners how you will provide feedback. Introduce any rubrics before starting the assignment and explain how you plan to review and grade assignments; let your learners know that you extend time and effort in doing so because you want them to develop skills and knowledge and reach course learning goals.

Incorporate more low-stakes formative assessment and lower the focus on higher-stakes summative assessment.

Students are more likely to cheat when they believe they need to cheat to pass or succeed.

Assessing early and often with lower-stakes assignments or activities, and spreading the course grade weight so that there are not as many high stakes exams, can lessen the pressure and provide more opportunities for learning. This helps students feel more prepared and confident in their ability to do the work themselves.

Students are also more likely to cheat if they’re focused on performance and outcomes, rather than mastery of skills and course topics.

There are many ways to incorporate assignments that encourage a focus on process and mastery. Any one of these strategies can have a positive impact:

• Scaffold major writing assignments or projects and provide formative feedback on activities such as idea generation, annotated bibliographies, and drafts; engage in process or application exercises in class. Consider focusing your feedback on these process steps; for end products or final assignments, just give a grade without extended comments. This can be a better use of your time, since learners may not review detailed comments once all is said and done.
• Use ungraded, relatively simple Classroom Assessment Techniques (CATs) to formatively assess student learning in real time. This page explains how to implement some useful CATs, including application cards, minute papers, and the muddiest point. These activities enable instructors to make informed instructional decisions (e.g., Should I review this apparent bottleneck?), and enable learners to make informed decisions about how to approach their learning (e.g., I thought I understood this until I tried to apply it; I need to practice this more before the exam!). Please reach out to the SAS teaching and learning team to discuss implementing CATs in your classes, in any modality and any class size. (Yes, CATs can scale up to very large classes!)
• Add “wraparound” practices around major assignments to focus students on process, skills, and plans for improvement. Before major assignments, discuss the assignment in a way that normalizes productive struggle as a part of the learning process. After major assignments, encourage metacognition and reflection: with exam wrappers or meta-assignments to focus students on process, skills, and plans for improvement:
  • For exams, use exam wrappers: brief reflective exercises that ask students about how they prepared for an exam, what kinds of errors they made, and what they could do differently next time.
  • For written assignments, use meta-assignments: ask students to write a brief paraphrase of your feedback on their paper, as they understand it.