Artificial Intelligence and Financial Reporting Quality in the Scientific Literature: A Bibliometric and Semantic Analysis

This study examines the intellectual and semantic organization of research on artificial intelligence (AI) and financial reporting quality (FRQ). It assesses the degree of semantic correspondence between the two domains. The analysis covers 311 peer-reviewed articles published between 2010 and 2026. The articles were identified through Scopus and Web of Science following PRISMA guidelines. Bibliometric mapping is combined with semantic text analysis using TF-IDF, cosine similarity, and semantic anchors. The results show rapid growth in scientific production and distinct thematic structures. There is no robust association between AI and FRQ semantic engagement: it is negligible with TF-IDF and with a second dictionary, and it is small with embeddings, positive with static word vectors and negative with a sentence encoder. Bibliographic coupling reveals shared intellectual foundations but limited semantic correspondence. Structural connectivity does not necessarily imply strong conceptual integration. The study distinguishes intellectual proximity from conceptual integration and identifies four potential mechanisms linking AI and FRQ: monitoring, information processing, organizational adoption, and external communication. At the dimension level, AI text-analytics research is positively associated with disclosure research.