FROM LEARNING TO EARNING: LONGITUDINAL INSIGHTS INTO HUNGARIAN GEN Z STUDENTS’ USE OF GENERATIVE AI IN ACADEMIC AND BUSINESS CONTEXTS (2023-2025)

The emergence of generative AI technologies—particularly large language models (LLMs) like ChatGPT—has introduced profound shifts in the academic and entrepreneurial practices of Generation Z students worldwide. This longitudinal mixed-methods study investigates the evolution of LLM adoption, perception, and utility among Hungarian university students between 2023 and 2025. The research builds on a systematic literature review of 34 peer-reviewed studies conducted according to PRISMA guidelines, which identified key global trends, benefits, and concerns surrounding LLM use in higher education and early-stage business contexts. Drawing from these insights, two large-scale national surveys were conducted in Hungary (n=442 in 2023; n=1,328 in 2025), with propensity score matching (PSM) applied to ensure comparability across cohorts. The findings reveal statistically significant increases in LLM usage frequency, perceived academic effectiveness, cognitive evaluations, and entrepreneurial innovation. These shifts reflect a growing strategic integration of LLMs into students’ learning routines and business activities, particularly among those with prior work experience or urban university affiliation. This study offers novel empirical insights into a regional higher education context where digital innovation uptake has traditionally lagged behind Western peers. The results indicate that how the literature see LLMs and AI technology, and Hungarian Gen Z students are rapidly closing the gap, moving from experimental to mainstream use of generative AI tools.