In A New Study, 94% Of AI-Generated College Exams Went Unnoticed And Consistently Outperformed Real-Life Students
Now that artificial intelligence (AI) platforms are widely accessible – with free models available online and AI-driven tools being integrated into social media apps – everyday people are beginning to use them for habitual tasks.
This includes students who are turning to AI for help with everything from English homework and science labs to exams.
The shift has prompted many universities to begin checking for AI-generated content in student work.
But, in a shocking new study conducted by researchers at the University of Reading in England, it was found that nearly all AI-written work was able to fly under the radar.
For the research, AI-generated exam answers were secretly submitted along with those of actual students in five undergraduate psychology modules. In the end, a whopping 94% of the AI-created submissions went unnoticed by examiners, even though they were completely created by an AI system with no human edits.
Plus, the AI submissions didn’t just go undetected. Rather, they also consistently outperformed real-life students.
The AI-generated answers scored, on average, half a grade higher than real students’ work. And in certain instances, the AI advantage nearly reached a full grade higher.
These findings call into question the future of education in the wake of AI and suggest that universities will have to evolve to safeguard exam integrity and the value of college degrees.
“Many institutions have moved away from traditional exams to make assessment more inclusive. Our research shows it is of international importance to understand how AI will affect the integrity of educational assessments,” explained Peter Scarfe, the study’s co-author.
Gorodenkoff – stock.adobe.com – illustrative purposes only, not the actual people
“We won’t necessarily go back fully to handwritten exams, but the global education sector will need to evolve in the face of AI.”
The study’s implications are much more far-reaching than just college campuses, too. As college students in the era of AI begin to graduate and embark on their professional careers, employers might start to question how reliable the academic credentials of applicants really are.
Still, the researchers underscored how the results of the study shouldn’t be viewed as a criticism of current practices in education. Instead, they hope the findings are viewed as a call to action and believe the education sector needs to adapt to AI as a new reality – balancing how to leverage AI while preserving the value and integrity of learning.
“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 wide use of AI in other areas of life to prevent a crisis of trust across society,” said Etienne Roesch, the study’s co-author.
“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.”
To read the study’s complete findings, which have since been published in PLOS ONE, visit the link here.
Sign up for Chip Chick’s newsletter and get stories like this delivered to your inbox.
More About:Science