Abstract
The accounting profession has undergone profound transformations from manual ledger systems to sophisticated artificial intelligence applications. This review examines the evolutionary trajectory of accounting practices, with particular emphasis on recent developments in large language models (LLMs) and their implications for contemporary accounting work. The transition from paper-based ledgers to enterprise resource planning (ERP) systems marked the first major digital transformation, fundamentally altering how financial data is recorded, processed, and reported. The subsequent integration of robotic process automation (RPA) automated routine tasks, while machine learning (ML) algorithms enhanced predictive capabilities in areas such as fraud detection and financial forecasting. Most recently, LLMs have emerged as transformative tools capable of processing natural language accounting queries, generating financial narratives, and supporting complex decision-making processes. This paper synthesizes current research on the application of artificial intelligence (AI) technologies in accounting, analyzing their impact on audit quality, tax compliance, financial reporting, and professional judgment. The review identifies key challenges including data privacy concerns, algorithmic transparency, regulatory compliance, and the evolving skill requirements for accounting professionals. Through systematic analysis of contemporary literature, this paper demonstrates how LLMs represent not merely incremental improvements but a paradigmatic shift in accounting practice, enabling natural language interfaces for financial systems and sophisticated analytical capabilities previously unattainable. The findings suggest that successful integration of LLMs requires careful attention to ethical frameworks, professional standards, and continuous adaptation of educational curricula to prepare future accountants for technology-augmented practice environments.

This work is licensed under a Creative Commons Attribution 4.0 International License.
Copyright (c) 2025 Huijie Fan, Jiaying Chen (Author)