Some school leaders and faculty are adamant about stamping out the smallest whiff of ChatGPT while others are prepared to sail into an AI revolution. However, new data suggests that students are turning in fewer AI-generated assignments and they are just as concerned about AI as you may be.
Copyleaks, an AI-based company known for its flagship plagiarism detection platform, found in its March report that student assignments found to have used AI dropped by 15%. Fewer than 10% of college assignments contained AI. When looking at specific types of colleges, the detection of artificial intelligence dropped at each one since February. Community colleges experienced the highest decrease, down 19.2% from February for just a total of 9.7% of assignments being detected to contain AI. Four-year institutions experienced the second-highest drop at 15.7%, down to a total of 9.9% of assignments.
Figures for other college types include:
- Private
- AI-detected assignments: 7.5%
- Decreased by 15.7% from February
- Career and Technical
- AI-detected assignments: 20.7%
- Decreased by .48% from February
Worldwide, across all college types, AI use also dropped.
Copyleaks began collecting anonymized data from high school and college students worldwide in January. Their multi-language AI Content Detector boasts a 99.2% detection accuracy rate. This figure extends to OpenAI’s recent unveiling of ChatGPT-4 as well.
Students may be submitting fewer assignments with the assistance of ChatGPT and related artificial intelligence tools due to recent innovations from plagiarism detector services that have caught up with the technology, such as Turnitin and Copyleaks’ AI Content Detector.
While student reliance on artificial intelligence tools has decreased this past month, College Rover discovered that 61% of high school students they surveyed who are seeking postsecondary admission are interested in attending a university that has banned the use of ChatGPT and other artificial intelligence tools. Men were particularly partial to this question, with 61% of them responding they’d prefer a university whose stance is strict on the technology.
Students are concerned about the ethical implications of using artificial intelligence, according to a similar report by BestColleges that surveyed college students. Specifically, 51% agreed that using artificial intelligence to accomplish assignments constituted cheating. Moreover, 41% agreed that it is morally wrong, compared to just 27% who disagreed.
Similarly, students may also have grievances over artificial intelligence’s ability to acquire race and gender biases. For example, one high school student found that it may contribute to systemic racism in America.
“I think that with ChatGPT, everything has its benefits and its downsides,” said one student, Ayala, according to USA Today.
More from UB: TikTok tracker: What colleges—or entire states—have banned the popular app?
