MAINTAINING ACADEMIC INTEGRITY THROUGH AI-TRAP QUESTIONS

Abstract

The current work introduces the concept of AI-trap questions as a tool for maintaining academic integrity in online courses. AI-trap questions are assessment tools designed to detect cheating by exploit-ing generative AI’s tendency to provide answers that are common rather than context-specific. This paper explores theoretical perspectives of academic integrity enforcement, examines the prevalence of cheating in higher education, and discusses the challenges posed by generative AI. The author proposes guidelines for developing and implementing AI-trap questions, highlighting their utility over alternatives, such as subjective assessments and current AI-detection software. While not a comprehensive solution, AI-trap questions may offer some instructors a tool to identify and deter AI-assisted cheating, particularly when integrated into a broader strategy for promoting academic honesty.