Quantifying Shortfalls in Students? AI-Supported Programming Practices

Generative AI assistants are permeating programming classrooms, yet little is known about whether students adopt sound usage habits.This study surveyed 50 undergraduates in Computer Science / Computer Information Systems courses to examine the frequency of 11 recommended AI-support behaviors, such as planning queries, verifying AI-generated output, and integrating AI with personal code.Substantial shortfalls were observed in critical forethought and verification steps, especially among students using AI for simpler tasks.In contrast, students applying AI to more complex assignments demonstrated stronger behaviors, although certain foundational habits remained weak across the board.The paper proposes five low-overhead scaffolds to help instructors close these usage gaps and improve AI literacy in programming education.