Recently, students of Texas A&M University got a shocking surprise due to a professor accusing them of cheating and taking disciplinary actions. The professor reportedly used ChatGPT, a generative AI platform, to detect if the submitted papers were AI generated. It is not uncommon for students to be tempted to cheat by using these AI tools, though it is not as effective in catching AI-generated work.
The professor’s false accusations resulted in 15 students being given an incomplete grade on their papers and had their diplomas placed on temporary hold pending an investigation. What the accused students didn’t know is that the professor fed their papers into the same ChatGPT platform to detect if the works were AI-generated. Sadly, the credentials of ChatGPT for verification are highly questionable since it also identified the professor’s 2021 research paper as AI-generated.
There are services that focus on detecting AI works but with little success, like Turn It In and GPTZero. According to reports, the professor in question didn’t make use of them either. Such blunders point to the possibility of ChatGPT being prone to misinformation and what some people refer to as “hallucinations”.
As a result of this unfortunate incident, it is important to note that AI platforms such as ChatGPT, Bing Chat, and Google Bard should not be used as the sole determination for whether a paper is artificially generated or not. It is also critical for students and professors alike to check the reliability, accuracy, and efficacy of such tools before any disciplinary decisions are made.
