<p>This study investigates the relationship between academic performance and students’ intent to use generative AI (GAI) for academic purposes, identifying a nonlinear, inverted U-shaped pattern where students with moderate academic performance exhibit the highest intent to use GAI. Drawing on the Yerkes-Dodson Law, this study suggests that moderate performers perceive GAI as a valuable learning aid, while high performers may see it as redundant, and low performers may lack the confidence or motivation to engage with it. Furthermore, this study examines the moderating roles of familiarity with GAI and academic integrity. Although familiarity with GAI did not significantly affect usage intent, academic integrity emerged as a crucial moderating factor. Specifically, lower-performing students with high academic integrity demonstrated greater intent to use GAI, viewing it as an ethical support tool for academic improvement. The study employs a quantitative survey methodology, collecting responses from 379 university students and analyzing the data using ANOVA and moderation analysis. These findings contribute to the broader discourse on technology adoption in educational settings, highlighting the ethical and motivational dimensions influencing students’ engagement with AI tools. The study provides insights for educators and policymakers to develop strategies that ensure responsible and effective AI integration in academia.</p>