This post may contain affiliate links, meaning at no additional cost to you, if you click a link and make a purchase, I will make a commission. Please read my disclosure for details.

Why an AI Classroom Policy?

With large language models like ChatGPT playing a growing role in our lives, professors are trying to adapt and develop classroom policies for the acceptable use of AI in projects and assignments. While one approach is simply to ban the use of AI in student work, the practicality of policing A.I. is difficult and fraught with challenges. Professors have found themselves in an unwinnable cat and mouse game. But the game is shifting. In this post, I’ll share the AI classroom policy I implemented this semester to try and guide and manage student AI use opposed to banning it.

Are AI Classroom Policies Practical?

Before we discuss policies, let’s put some things out on the table. 

First, while I have a number of personal and ethical concerns about AI, I also know that it is my professional responsibility as an educator to try to guide students in their use of it. The reality is that I can’t stop students from using it in their lives just as I can’t stop the wave that is AI. 

Second, it can be very difficult at times to discern if a student is using AI in a way that violates the classroom policy. 

Third, there are likely going to be some students who try to get around your AI policy, whatever it is. But I don’t think that’s the case for most students. I think most students want to know what the expectations are so that they can use artificial intelligence as a tool to enhance their work. 

The first time I taught my writing class following the ChatGPT explosion, I worried that students would simply plug the assignment into ChatGPT and turn in what it spits out. But students told me they didn’t want AI to replace their minds or their creative outputs. Many of them actually seemed more apprehensive about adopting AI in the classroom than I was. 

Since then, we’ve all grown more familiar, comfortable, and savvy. As we transition to a more mature and nuanced use of these tools, we can guide our students toward ways they can use these tools beyond the low-hanging fruit of ‘write this paper for me.’ 

Developing an AI classroom policy will help your students by establishing clear expectations, helping to stave off unnecessary issues and conflicts. 

The New AI Policy in my Classes

My AI policy on assignments has tended to vary, making it confusing for students. For example, in my writing class, in one assignment I let students use a large language model to brainstorm ideas. On a different assignment, I changed it up, carving out a new area where LLMs were allowed. This sort of made sense for the unique nature of the writing assignments in that class. 

But as I tried to figure out how I wanted to address AI in other classes, I realized my piecemeal approach wouldn’t bode well for larger tasks, like research projects. 

I found myself stuck trying to think of an extensive list of what was and wasn’t allowed. 

I finally got the clarity I needed when I attended Shepherd University’s Faculty Professional Development Day at the start of the semester. I had the opportunity to attend the panel “Redesigning Assessment in the Age of AI” by Instructional Designer Yildiz Nuredinoski and Professor of Biology Dr. Sher Hendrickson. In the presentation, Dr. Hendrickson provided a policy adapted and tweaked from a policy shared by Marc Watkins in his course Generative AI in Teaching.  With their permission, I share the policy as I’ve implemented it below, with a few minor tweaks of my own. 

Acceptable Use of AI includes:

• Brainstorming ideas
• Exploring topic ideas in depth.
• Using AI-assisted research tools such as SciSpace, ExplainPaper, or Perplexity to search for sources.
• Using AI tools to search for information within a paper (such as with SciSpace or NotebookLM). [In this post, I went over using NotebookLM to search within a document].
• Exploring counterarguments and viewpoints using AI (such as with Perplexity).
• Using suggestions from AI tools such as Grammarly to improve the clarity of conciseness of your writing.

Note that students are expected to cite their use of AI.

Not Acceptable Use of AI

In my classes, the below uses of AI would result in an F on the assignment.

• Prompting an LMS with the assignment and using the results as your work.
• Using AI to generate text as a substitute to your own research, reasoning, and thoughts.
• Using AI to write large chunks of text of your paper with little or no input by you. If more than 20% of a paper is identified as being written by AI, it will result in an F.
• Trusting something that AI has generated at face value without verifying sources.

Additional Classroom AI Policy Resources

A number of great resources have emerged for developing the AI classroom policy that works best for your teaching style, subject matter, and classroom.  For example, Stanford’s Teaching Commons offers a pretty thorough guide for creating your course AI policy. CSU Bakersfield has collected several policy examples. Lastly, here are some reasons why faculty have molded their AI class policies as they see necessary.

Conclusion

Thank you to Marc Watkins and Dr. Hendrickson for allowing me to share this policy and how I’m implementing it in my classes on this blog. I hope this post helps you as you think about your own AI classroom policy. 

As I like to do, I made a few videos to support some of the ways in which AI could be used in alignment with the policy for a project. I’ll share those videos in the next blog post. 

Be well! 

-Matt