LARGE LANGUAGE MODELS AND THE TRANSFORMATION OF PROFESSIONAL ACCOUNTING PRACTICE: FROM DATA PLUMBING TO STRATEGIC INTERPRETATION

Large language models (LLMs) are reshaping the professional landscape of accounting and finance, shifting practitioners from labor-intensive data preparation toward higher-order analytical reasoning. This paper examines how LLMs can be strategically deployed across core accounting workflows - from financial document analysis and regulatory compliance to audit risk assessment - while identifying the governance structures necessary to ensure professional integrity. Drawing on recent empirical benchmarks, industry case studies, and the emerging retrieval-augmented generation (RAG) architecture, we argue that the critical variable is not whether LLMs are used, but whether they are deployed within a framework that preserves professional judgment, data privacy, and regulatory fidelity. Implications for accounting educators, practitioners, and firms undertaking phased AI adoption are discussed.