Abstract: Student use of generative AI tools, such as ChatGPT, is a growing concern in all aspects of higher education. However, research on the impact of these tools on student learning is currently limited, especially in more technical programs such as information systems. The research that does exist yields mixed results and indicates that the impact on performance may be dependent on the course type. Universities and faculty have been left to struggle with the development of policies and procedures in the face of a technology that is rapidly becoming ubiquitous. This paper documents a study of students in a technical database management systems course, where students were allowed but discouraged to use AI tools to aid in the completion of assignments. Student use of AI was tracked throughout the semester. The results indicate that the use of AI is predicted by performance earlier in the semester. AI users did not perform at lower levels than non-users on technical assignments but did perform at a lower level than their non-AI using colleagues on the final exam, where the tools were unavailable. AI users were also found to perform consistently over time. These results shed light on the effects of unrecommended AI use and indicate that faculty need to carefully incorporate AI tools and policies into their more technical courses.

Keywords: Generative AI, Generative AI in teaching, Structured query language (SQL), Learning improvement, Student learning

Download This Article: JISE2026v37n4pp526-545.pdf

Recommended Citation: Collignon, S. E., Peace, A. G., Peace, L., & Surendra, N. C. (2026). Unrecommended Use of Generative AI: Decision to Use and Effects on Performance and Learning. Journal of Information Systems Education, 37(4), 526-545. https://doi.org/10.62273/SGYZ6748