AI or Human? Finding and Responding to Artificial Intelligence in Student Work

Abstract

IntroductionRecent innovations in generative artificial intelligence (AI) technologies have led to an educational environment in which human authorship cannot be assumed, thereby posing a significant challenge to upholding academic integrity.Statement of the problemBoth humans and AI detection technologies have difficulty distinguishing between AI-generated vs. human-authored text. This weakness raises a significant possibility of false positive errors: human-authored writing incorrectly judged as AI-generated.Literature reviewAI detection methodology, whether machine or human-based, is based on writing style characteristics. Empirical evidence demonstrates that AI detection technologies are more sensitive to AI-generated text than human judges, yet a positive finding from these technologies cannot provide absolute certainty of AI plagiarism.Teaching implicationsGiven the uncertainty of detecting AI, a forgiving, pro-growth response to AI academic integrity cases is recommended, such as revise and resubmit decisions.ConclusionFaculty should cautiously embrace the use of AI detection technologies with the understanding that false positive errors will occasionally occur. This use is ethical provided that the responses to problematic cases are approached with the goal of educational growth rather than punishment.