Two years after ChatGPT’s release, its disruptive influence is starkly evident in education. Students are increasingly using generative AI for assignments and exams, submitting chatbot-generated work as their own to earn grades, credits, and even degrees.
New research from the U.K. reinforces a critical concern: teachers are largely unable to detect AI-generated academic work.
The increasing use of AI to complete academic work poses a significant threat to the value of high school diplomas and college degrees. It also raises serious concerns about unqualified individuals entering critical professions like nursing, engineering, and firefighting, where a lack of genuine knowledge could have dire consequences.
Alarmingly, some schools have exacerbated the problem by permitting AI usage while prohibiting the very technology designed to detect it, effectively encouraging academic dishonesty.
A study by Peter Scarfe and colleagues at the University of Reading, U.K., reveals a concerning inability of educators to identify AI-generated work. Researchers submitted basic AI-generated assignments under fake student profiles, and a staggering 97% went undetected. Even more troubling, the report suggests this 6% detection rate is likely inflated compared to real-world cheating scenarios.
Here is what the report said:
Overall, AI submissions verged on being undetectable, with 94% not being detected. If we adopt a stricter criterion for “detection” with a need for the flag to mention AI specifically, 97% of AI submissions were undetected.
The difficulty in identifying AI-generated text isn’t new. A University of South Florida study last year demonstrated that even trained linguists struggled to differentiate between AI-created content and human writing.