GENERATIVE ARTIFICIAL INTELLIGENCE IN VOCATIONAL UNDERGRADUATE FINANCE AND ACCOUNTING EDUCATION: APPLICATIONS, RISKS, AND FUTURE DIRECTIONS

Abstract

Generative artificial intelligence is entering finance and accounting classrooms before curricula, assessment systems, and institutional controls have fully adapted. This narrative review explains the technical principles of large language models and examines their use in vocational undergraduate finance and accounting education. Potential applications include dialogic tutoring, formative feedback, text simplification, case simulation, spreadsheet and programming support, financial analysis, professional communication, and course design. These functions can make practice more accessible and provide rapid opportunities for comparison and revision. Their educational value, however, depends on how the task is structured. Fluent output may contain fabricated facts, obsolete rules, incorrect calculations, biased assumptions, or unsupported recommendations. Unrestricted assistance may also improve immediate task performance without developing the independent knowledge and professional judgment that students need. Additional concerns include academic misconduct, confidential-data exposure, unequal access, unreliable detection, and weak accountability. A responsible integration model should therefore separate AI-free foundations, guided AI assistance, and AI-critical professional tasks. It should require authoritative-source verification, calculation checks, process records, disclosure, reflection, and oral or practical defense. Future systems may combine domain-controlled retrieval, multimodal simulations, and auditable learning records, but vocational programmes need direct longitudinal and multi-institutional evidence. Generative AI should be treated as a supervised learning partner and an object of professional evaluation, not as an authoritative answer engine.