How to Use AI to Detect Cheating in School

In 2024, cheating has evolved. It’s a story as old as teaching: A student, for some reason, uses someone else’s work to finish an assignment. But now he could be using an artificial intelligence tool.

The appeal is understandable. Away with those dodgy essay writing services where a student has to pay real money for some unscrupulous person to write 1200 words about the fall of the Roman Empire. An AI writing tool can do it for free in 30 seconds.

As a professor of strategic communications, I regularly encounter students using AI tools like ChatGPT, Grammarly, and EssayGenius. It’s usually easy to tell when a student has used one of these tools to compose their entire paper. The telltale signs are ambiguous language and a super-annoying tendency for the AI ​​to spit out text with the command prompt plastered all over it.

For example, a student can use ChatGPT, an AI tool that uses large-scale language modeling techniques and a conversational question-and-answer format to deliver search results. This allows them to write a short essay in response to an assignment by simply copying and pasting the essay question into the tool.

Follow this assignment: explain in 300 words or less how this SWAT and brand audit will impact your final pitch.

This is the result of ChatGPT:

I have had this type of response, or responses very close to it, a few times in my time as a teacher. One of the most recognizable warning signs is the number of times key terms from the assignment are used in the final product.

Normally, students do not repeat the key concepts from the assignment in this way in their paper. The results therefore look more like old-fashioned SEO texts that define these concepts than a unique essay that must demonstrate insight into the subject.

But can teachers use AI tools to catch students using AI tools? I’ve come up with a few ways to be smarter about spotting artificial intelligence in papers.

Catching cheaters with AI

Here’s how you can use AI tools to catch cheaters in your classroom:

1. Understand the possibilities of AI: There are now AI tools on the market that can scan an assignment and its assessment criteria to deliver a fully written, cited, and complete piece of work in moments. Familiarizing yourself with these tools is the first step in the fight against AI-driven integrity violations.
2. Do what the cheaters do: Copy and paste all of your assignments into a tool like ChatGPT before the semester starts and ask it to do the work for you. When you have a sample of the type of results it specifically produces in response to your assignments, you’ll be better equipped to spot robot-written answers. You can also use a tool specifically designed to spot AI writing in papers.
3. Get a real sample of your text: At the beginning of the semester, ask your students to submit a simple, fun, and personal piece of writing to you. The assignment should be something like “200 words about what your favorite toy was as a kid” or “Tell me a story about the most fun you’ve ever had.” Once you have a sample of the student’s real writing, you can later use it to have an AI tool compare that sample to what you suspect is AI-written work.
4. Request a rewrite: If you suspect a student is using AI to cheat on their assignment, take the submitted work and ask an AI tool to rewrite it for you. In most cases I’ve encountered, an AI tool will rewrite its own work in the laziest way possible, substituting synonyms rather than changing any material elements of the “original” work.

Here’s an example:

Now let’s look at what a real human being (me) wrote: my CNET bio:

The wording has been changed, taking much of the soul out of the writing and replacing it with sentences that are arguably clearer and more direct. There are also more additions to the writing, presumably for clarity.

The most important part of catching cheaters who are using AI to do their job is having a reasonable amount of evidence to show the student and the administration at your school when it comes down to it. Maintaining a skeptical attitude when grading is vital, and your ability to demonstrate ease of use and understanding with these tools will make your case even stronger.

Good luck out there in the new AI frontier, fellow teachers, and try not to be offended when a student turns in work written by their robot colleague. It’s up to us to make the prospect of learning more appealing than the temptation to cheat.