AI Disclosure as an Institutional Investment Catalyst: Evidence from China

Abstract Artificial intelligence (AI) is a significant driver propelling the new wave of technological revolution and industrial transformation, exerting a profound influence across various sectors. This paper examines the relationship between AI information disclosure and changes in institutional investor shareholdings, utilizing data from Chinese A-share companies listed on the Shanghai and Shenzhen stock exchanges between 2007 and 2023. The results indicate that enhanced AI information disclosure is associated with increased institutional investor ownership. This conclusion remains valid following robustness tests that include substituting dependent and independent variables, altering clustering levels, excluding municipal samples, and employing quantile regression. Heterogeneity analyses further reveal that the effect is notably stronger in samples post-2011, within manufacturing firms, in environments with higher competition, and among firms with directors possessing a strong technological background. The mechanism analysis reveals that increased AI information disclosure prompts institutional investors to conduct on-site visits, enhancing information transparency and bolstering innovation capabilities. Further, heightened AI information disclosure is particularly significant in encouraging shareholdings by long-term and resilient institutional investors. This paper presents the first micro-level evidence on the impact of AI disclosure on institutional investor shareholdings, offering theoretical and practical implications for refining investment strategies of institutional investors in the capital market.