Detecting AI-Generated Content

As AI writing tools become more prevalent, it's becoming increasingly difficult for teachers to determine whether a student's work is genuine or generated by a machine. A professor of strategic communications has encountered students using AI tools like ChatGPT, Grammarly, and EssayGenius on a regular basis. To combat this, the professor has developed strategies to detect AI-generated content, including analyzing language and using AI tools to spot cheating.

One tell-tale sign of AI-generated content is the use of ambiguous language and the tendency for AI to repeat key terms from the prompt. The professor notes that students don't normally repeat key terms from the prompt in their work in this way, and the results read closer to old-school SEO-driven copy meant to define these terms rather than a unique essay meant to demonstrate an understanding of subject matter.

To catch AI cheaters, the professor suggests using AI tools to scan an assignment and its grading criteria to provide a fully written, cited, and complete piece of work. This can help teachers familiarize themselves with the capabilities of AI tools and spot potential cheating. Additionally, requiring students to submit a simple, fun, and personal piece of writing at the beginning of the semester can provide a sample of their real writing style, which can be used later to compare with suspected AI-generated work.

The professor also recommends using AI tools to rewrite suspected AI-generated work. In most cases, an AI tool will rewrite its own work in a lazy manner, substituting synonyms instead of changing any material elements of the original work. By maintaining a skeptical mind when grading and demonstrating ease of use and understanding with these tools, teachers can make a stronger case against cheating and promote academic integrity.