Search for a command to run...
This study investigates how students perceive mobile, generative AI tools such as ChatGPT as regulating cognitive load and emotional demands during learning. The research examines how AI-mediated mobile communication affects intrinsic, extraneous, and germane cognitive load, and how these perceptions influence mental effort, cognitive engagement, learning anxiety, and perceived academic performance. The study is situated in Nepal, where mobile learning and smartphone-based educational practices often substitute for limited institutional support. Guided by Cognitive Load Theory, a structural equation model was developed to examine relationships between perceived AI-mediated load types and student outcomes. The model was tested using survey data from 314 higher education students who had experience using ChatGPT through mobile devices. Results show that students who perceived ChatGPT as reducing extraneous load and enhancing germane load reported lower anxiety and higher cognitive engagement. These perceptions were also associated with higher perceived academic performance and more efficient mental effort. Intrinsic load remained a significant constraint but was mitigated when ChatGPT was used as a mobile scaffold in complex learning tasks. This is one of the first theory-driven studies to empirically test how students perceive generative AI's influence on multiple dimensions of cognitive load in mobile learning environments. The research contributes theoretical and empirical insights to information systems and educational technology by reframing mobile AI not merely as a usability tool but as a mechanism for cognitive load management and emotional regulation in under-resourced educational systems. The findings offer practical guidance for educators and instructional designers seeking to integrate generative AI into mobile learning strategies. ChatGPT can serve as a portable cognitive and emotional support system that clarifies complex concepts, reduces learning anxiety, and enhances engagement. Incorporating guided mobile AI use into coursework can improve learning efficiency and emotional well-being without increasing instructional workload, particularly in resource-constrained higher education settings.