Cheating with ChatGPT and Techniques of Neutralization

This paper utilizes neutralization theory to understand cheating with AI-assisted programs among university students. Using data collected from undergraduate students across the United States, we explore how criminological concepts and demographics predict the number of times someone cheated over the past year as well as the likelihood of using AI-assisted tools to cheat. We find agreement with cheating neutralization techniques, general cyber-neutralization techniques, and the presence of deviant peers positively associate with all cheating outcomes, while sense of guilt is inversely associated. Low self-control is positively associated with use of AI-assisted misconduct, but not the number of times students cheated.