Artificial intelligence innovation and financial report quality

This paper examines the effect of artificial intelligence adoption on firms' annual report disclosure quality. I employ a text analysis approach and use MD&A section texts to measure the precision and readability of corporate annual reports. Using a panel dataset of Chinese listed firms from 2010 to 2023, I find that corporate AI adoption effectively enhances the readability and accuracy of text in the MD&A section. Specifically, corporate AI adoption could increase management efficiency and enhance internal control over information processing which reflects its governance effect on disclosure. Furthermore, the main effect of corporate AI adoption was more pronounced when AI patents are more closely related to firms' core business and when firms’ monopoly power is relatively weak. The above conclusions remain robust after employing the DID model and instrumental variable approach. I provide further evidence of the theoretical link between corporate AI adoption and disclosure behaviour, while extending the literature on determinants of management disclosure and AI outcomes in capital markets. • Artificial intelligence adoption will impact firms' disclosure quality positively. • AI technology will improve efficiency and effectiveness on information processing. • AI-Business relevance reveals the differential importance of AI innovation and its various effects on disclosure activities. • Supervised machine learning methods can uncover more in-depth information from patent data.