An Interpretivist Socio-Technical Research Framework for Understanding Antecedents for the integration of Artificial Intelligence in Financial Reporting

Abstract

Artificial intelligence (AI) is increasingly entering accounting and financial reporting through automation, machine learning, natural language processing, predictive analytics and generative AI. This expansion has generated a rapidly growing literature concerned with adoption, acceptance, implementation, performance and the consequences of human–AI interaction. Nevertheless, much of the literature continues to represent AI integration through technology-adoption models that privilege variables such as perceived usefulness, behavioural intention, technological readiness or organisational determinants. These approaches provide valuable explanations of adoption, but they can provide limited methodological guidance for understanding how financial professionals interpret AI through their prior technological histories, professional development and organisational circumstances. This paper develops an interpretivist socio-technical research framework for investigating antecedents of AI integration in financial reporting. Drawing on interpretivism in information systems research and Socio-Technical Systems Theory, the framework conceptualizes AI integration as a socially interpreted, historically situated, organizationally embedded, and recursively evolving process rather than as a discrete adoption event. Four antecedent domains are integrated: financial professionals’ perceptions of AI; prior experience with traditional computer-based accounting and reporting systems; professional skills development; and organizational contextual factors. The paper makes a methodological contribution by translating these domains into a research architecture specifying how each can be investigated through semi-structured interviews, observation, and document analysis. It further develops an interpretive analytical cycle linking experience, interpretation, interaction, organizational response, and adaptation, and establishes methodological principles concerning contextuality, historical experience, socio-technical interdependence, practice orientation, triangulation, reflexivity, and emergence. The framework is intended to guide qualitative and mixed-method research on AI-enabled financial reporting and is particularly relevant to organizational contexts in which technological transformation remains intertwined with professional capability, legacy systems, institutional expectations, and resource constraints.