The growing disconnect between firms’ symbolic AI disclosure and their substantive AI investment, known as AI washing, threatens disclosure quality and capital allocation efficiency. Using Chinese A-share listed firms from 2017 to 2024, we examine whether institutional investors discipline corporate AI washing. Institutional ownership significantly reduces AI washing, an effect driven by long-term institutions while short-term institutions play no role. Mechanism analyses identify two complementary channels: a monitoring channel that improves information transparency and constrains managerial impression management, and a resource channel that promotes substantive AI R&D. The effect is stronger among non-state-owned firms, firms with lower analyst coverage, and during the AI hype period, and is concentrated among independent institutions. We further show that AI washing significantly depresses medium- to long-term firm value, providing an economic rationale for institutional governance engagement. We offer systematic evidence on curbing AI washing from a governance perspective.
Ye et al. (Wed,) studied this question.