Does Writing About AI Detection Tools Benefit Ethical Conduct?

Work thumb

Views: 27

All Rights Reserved

Copyright © 2025, Common Ground Research Networks, All Rights Reserved

Abstract

This study examined whether an assignment devoted to the accuracy of artificial intelligence (AI) detection programs would deter first-year students from relying on generative AI writing tools. Students who were enrolled in a written communication course dedicated to scientific writing first conducted and then wrote about a study on cognitive biases (control condition) or a study on the accuracy of AI detection programs (experimental condition). All students were second-language speakers of English who perceived the course as challenging. Thus, the enticement of using generative AI writing tools to aid writing was considerable despite explicit prohibition. During class meetings, students wrote the introductory section of their research reports and received feedback from the instructor. Then, they revised their work at home before submission. The topic of the assignment did not deter students from relying on generative AI writing tools. On the contrary, students in the experimental condition were more likely to use such tools. Furthermore, the more students relied on such tools, the lower their grades on the assignment as scored by independent raters unaware of each student’s classification as either AI user or non-user. These findings suggest that students’ informed awareness of the likelihood of detection may not yield the expected educational outcomes.