Challenges and Responses to the Misuse of Artificial Intelligence in Chemistry Education in China
TL;DR
Evidence indicates that moral regulation and technological surveillance alone are inadequate to protect academic outcomes from AI intrusion, and highlights the urgent need to redesign assessments to be AI-resilient, emphasizing process over final answers, and fostering ethical awareness.
Abstract
The rapid integration of Generative Artificial Intelligence (AI) into higher education presents a dual reality: it functions as a powerful pedagogical tool while simultaneously creating unprecedented challenges to academic integrity. This study examines the impact of AI on academic integrity in chemistry education. Focusing on second-year environmental engineering undergraduates at Jinggangshan University, online physical chemistry examinations (N = 104) were compared with a historical control group assessed through paper-based exams (N = 54). The results revealed significant score inflation in the online format, inconsistent with students’ prior performance in mathematics and general chemistry. Data from the smart examination system further indicated that 18.23% of students engaged in AI-assisted misconduct, employing techniques such as background switching and AI-embedded input methods to evade proctoring. Although the head teacher support model contributed to the development of more constructive learning habits, evidence indicates that moral regulation and technological surveillance alone are inadequate to protect academic outcomes from AI intrusion. Therefore, these findings highlight the urgent need to redesign assessments to be AI-resilient, emphasizing process over final answers, and fostering ethical awareness. The study advocates for a balanced framework of responsible AI use that preserves both pedagogical benefits and rigorous evaluation standards.