AI Chatbots in College Education: Navigating Academic Integrity and Critical Engagement

Dan Barnett, Community College Philosophy Instructor

Here’s the text that I append to every writing assignment in my online classes:

Explain how you formulated your response. 

What sources did you use (including your own thinking process, course materials, websites, and AI chatbots)? 

If you used material from a chatbot, how did you verify what it wrote? 

How did you use it—as a writing partner whose words you edited? 

Or are you quoting directly from what it wrote (if so, be sure to use quotation marks)?

****

POLICY ON THE USE OF ARTIFICIAL INTELLIGENCE (AI) CHATBOTS

The goal of this course is to enlarge your understanding of the world and provide you with an opportunity to make key concepts your own. Some of these concepts you may find to be life-changing.

But in order for you to make that determination, it’s important that you have first-hand experience with the material presented in this class. It’s like the difference between reading a Wikipedia article on basketball and actually learning how to play. The Wikipedia article may help you answer a quiz question, but making the game your own may affect the way you exercise, the way you spend your time, who you root for, even the friends you have.

But since this is a college class there are of course quiz questions (and other assignments) and limited time to get all the work done in all your classes. If time is short, and a milestone date is fast approaching, it’s only human to reach out to the internet or use an app for help. Google can get you to a site with the information or interpretation you need (and the search field can even do math problems!).

When such materials are copied and pasted into Canvas without attribution that is straightforward plagiarism (defined as falsely claiming the writing as one’s own). That's different from going to Google or an app to find a definition or other factual information or common knowledge or doing research and providing citations.

But now things have become much more complicated with the advent of freely available chatbots using AI. You can ask for summaries of difficult or long texts, interpretations of ideas, arguments pro and con on just about anything, and even creative expressions (like writing poetry). Unlike a link from a Google search, which takes you to someone else’s material, the chatbot produces new material for the occasion.

So is it plagiarism to use what chatbot produces? No, if the source is given. But most chatbots are not programmed to give you their source(s), which may be faulty. The chatbot tends to produce material that sounds authoritative, but it may be flat wrong.

That brings up a second issue related to the purpose of this class. If you ask a chatbot to write something for you, because the time is short, and you put the output in quotation marks and give the source as the chatbot and submit this as your answer—you haven’t really engaged with the concept. The chatbot may be flat-out inaccurate but you’ve been taken in by its authoritative presentation.

At the same time chatbots can be very helpful as a kind of writing partner, giving you words you can build from (thanks to their sophisticated algorithms) instead of a blank computer screen. A human tutor can do the same thing as you kick ideas back and forth and develop ways of expressing a concept.

For that reason it doesn’t make sense to ban the use of AI chatbots. In fact, as Google and other search engines integrate their own AI, and other apps integrate AI as well, you may be using it and not even know.

Years ago there was controversy over whether math classes should allow the use of calculators. It turned out those little devices didn’t kill off math classes but they did change the focus to “show your work.” Yes, getting a right answer was important, but how you arrived at that answer was the key to learning.

We face another such question: Will the use of AI chatbots end college writing? My answer is no, but the focus will change from the student just delivering “an answer” and moving on—to the submission not just of an answer but also information about how the answer was constructed, what sources were consulted (including course materials) and whether the writer specifically engaged an AI chatbot to help with the writing and, if so, did “source checking” to confirm what the chatbot presented.

The bottom line is for you not just to affirm responsibility for the answer submitted but to provide evidence showing why you think it’s correct (whether you get help from a chatbot or not). In fact, a healthy skepticism about what the chatbot writes can form a good critical thinking mindset on your part and spur engagement with the concepts you’re writing about.

If the availability of AI chatbots can usher in a new focus not just on the answer but how the answer came to be, including the evaluation of its sources, that’s the kind of engagement that we instructors dream about. You will be making key concepts your own.

I asked ChatGPT to summarize what I’ve written. Here’s the response:

“The goal of the course is to expand the student's understanding of the world by providing them with key concepts, which may be life-changing. However, it's important for students to have first-hand experience with the material presented in the class. The use of AI chatbots has become complicated, as it can produce new material that sounds authoritative, but may be incorrect. The use of chatbots should not be banned, as they can be helpful as a writing partner, but the focus will change from the student just delivering an answer to the submission of an answer along with information about how the answer was constructed, what sources were consulted, and whether the student specifically engaged an AI chatbot to help with the writing. The student should be responsible for the answer submitted and provide evidence to show why they think it's correct.”