AI May be Outsmarting Your Teacher

On average, fake answers generated by artificial intelligence earned higher grades than those achieved by real students.

07.02.24
AI May be Outsmarting Your Teacher (Getty Images)

Exam submissions generated by artificial intelligence went almost entirely undetected by markers at a leading university.

And, on average, fake answers generated by AI earned higher grades than those achieved by real students.

Researchers say their worrying findings, published in the journal PLOS One, suggest that experienced exam markers may “struggle” to spot AI-generated answers – and should act as a “wakeup call” to education chiefs.

In recent years, machine learning tools such as ChatGPT have become more advanced and widespread, leading to concerns about students using them to cheat by submitting AI-generated work as their own.

Such concerns are heightened by the fact that many universities and schools switched from supervised in-person exams to unsupervised take-home exams during the Covid pandemic, with many now continuing such models.

Tools for detecting AI-generated written text have so far not proven very successful.

To better understand the issues, a Reading University research team generated answers that were 100% written by the AI chatbot GPT-4 and submitted on behalf of 33 fake students to the examinations system of the University’s School of Psychology and Clinical Language Sciences.

The exam markers were unaware of the study.

The research team found that 94% of their AI submissions went undetected.

On average, the fake answers earned higher grades than real students’ answers.

In 83.4% of cases, the AI submissions received higher grades than a randomly selected group of the same number of submissions from real students.

The results suggest the possibility that students could not only get away with using AI to cheat, but they could also achieve better grades than achieved by their colleagues who do not cheat.

The research team says that a number of real students may have gotten away with AI-generated submissions in the course of the study.

 Updated advice has already been issued to university staff and students as a result of the findings of the largest and most robust study of its kind to date.  

Originally published by Talker News

Support the Next Generation of Content Creators
Invest in the diverse voices that will shape and lead the future of journalism and art.
donate now
Support the Next Generation of Content Creators
Invest in the diverse voices that will shape and lead the future of journalism and art.
donate now