Construction and Pathway Exploration of a “Lightweight” Generative AI Model for Fundamental Accounting Education: A Study Based on the TPACK Framework

Abstract

The digital intelligence transformation has profoundly reshaped the capability landscape of the accounting profession, necessitating an urgent paradigm shift in accounting education. However, current accounting instruction faces significant dilemmas, including an overreliance on “heavy-asset” specialized software, high barriers to technological adoption, and ineffective integration of core competencies. These challenges hinder the systematic development of students' ability to leverage intelligent technologies for solving complex accounting problems. To address these bottlenecks, this study draws upon an evolved TPACK framework to propose an innovative “lightweight integration” teaching paradigm. This approach advocates for the use of general-purpose generative AI as a low-threshold technological foundation, replacing expensive specialized educational systems. Through “minimally invasive” instructional design, it facilitates a dynamic and deep integration of Technological, Pedagogical, and Content Knowledge. The study constructs a comprehensive “Environment–Content–Process–Evaluation” integrated teaching model. Using the “Intelligent Audit of Original Vouchers and Risk Insight” module from Fundamental Accounting as a case study, it demonstrates the complete pedagogical chain from prompt design to “human-machine co-verification.” This model effectively lowers the economic and technical barriers to digital intelligence education, offering a theoretically self-consistent and cost-effective solution to the disconnect between technology and professional practice in accounting education reform.