Researchers at the University of Reading have found that artificial intelligence (AI)-generated exam answers have gone almost entirely undetected and consistently achieved higher grades than those written by real students.
The study involved submitting answers generated by ChatGPT-4 on behalf of 33 fictitious students to the School of Psychology and Clinical Language Sciences’ examination system.
“In terms of standard exams, I would expect similar results in most disciplines. AI systems such as ChatGPT have been trained on a huge corpus of data. Not just related to Psychology,” Scarfe told IE.
To read more, click here.