Impact of Excessive AI Tool Usage on the Cognitive Abilities of Undergraduate Students: A Mixed Method Study

This is a mixed-methods research that explores how the excessive usage of AI tools influences the cognitive skills of undergraduate students, including the perceived advantages and possible long-term outcomes. Using a sequential explanatory research design with quantitative survey (n=300) and qualitative interviews (n=45), the study exposes that there are highly relevant correlations between high AI dependency and lower critical thinking skills (17.3 percentage points lower scores) and worse memory retention (22 percent fewer concepts retailed). Although students noted that AI technologies were more efficient and accessible in terms of use, the qualitative data revealed problematic trends in cognitive offloading, a lack of motivation to engage in deep learning, and the beginning of a sort of solution paralysis when students cannot work without the help of technology. That fits with Cognitive Load Theory and Digital Dependency theories, which imply that unrestrained use of AI can cause a degradation of basic cognitive abilities due to decreased expenditure of mental effort. It is worth noting that differences across disciplines also occurred with humanities students recording the sharpest cognitive drops. One of the most important paradoxes of educational technologies that the study points out is that the tools that are being implemented to facilitate learning can damage equally the skills that they are trying to facilitate when being over-utilized. The findings have significant educational implications, arguing a middle-road approach to AI integration, retaining cognitive engagement and taking advantage of technological opportunities. The discussion provides three evidence-based concepts of a sustainable implementation, which are strategic scaffolding, metacognitive wrappers, and cognitive load calibration. Such limitations as the heterogeneity of the samples and the biases of self-reportingare mentioned, which can hint at further longitudinal and neurocognitive studies. The paper makes a contribution to an emerging body of research on the place of AI in education by offering an empirical demonstration of the cognitive tradeoffs associated with AI, and finally proposes pedagogically minded solutions to avoid the atrophy of critical mental abilities in the digital age.