This post is also available in:
עברית (Hebrew)
The largest blind study to date that challenged human educators to detect AI-generated content showed that the AI-generated answers went undetected in 94% of cases.
The study, conducted at the University of Reading in the UK, was meant to identify potential risks and opportunities of AI for research, teaching, learning, and assessment. After finding that even highly experienced exam markers struggled to spot AI-generated answers, the researchers are calling for the global education sector to do more to address this emerging issue and form the appropriate policies.
According to Techxplore, the researchers had ChatGPT generate exam answers for several undergraduate psychology modules. They then performed a blind test of a real-life university examinations system, where the AI-generated answers went undetected in 94% of cases and even got higher grades than the real student submissions.
Associate Professor Peter Scarfe and Professor Etienne Roesch said these findings should be a “wakeup call” for educators across the world, as a recent UNESCO survey of 450 schools and universities worldwide found that less than 10% of them had policies or any guidance regarding the use of AI.
Professor Roesch said: “As a sector, we need to agree on how we expect students to use and acknowledge the role of AI in their work. The same is true of the wider use of AI in other areas of life to prevent a crisis of trust across society… Our study highlights the responsibility we have as producers and consumers of information. We need to double down on our commitment to academic and research integrity.”
Pro-Vice-Chancellor for Education and Student Experience at the University of Reading, Professor Elizabeth McCrum, said that as AI is transforming our lives, this includes how we teach students and assess their learning. She explains that possible solutions include moving away from older methods of assessment and progressing to one more aligned with workplace skills, including the use of AI.
