Abstract:
Corporate AI-washing—shaping an image of technological advancement by exaggerating one's own AI capabilities in order to attract investor attention and improve corporate performance—may affect market expectations and corporate innovation. Using Shanghai and Shenzhen A-share listed companies as samples, this paper constructs an AI-washing indicator based on annual report texts and AI patent data, and finds that AI-washing significantly inhibits corporate innovation, as evidenced by significant reductions in both innovation quantity and innovation quality. However, neither the market attention nor the operating efficiency of enterprises improves. The underlying mechanisms are as follows: First, enterprises allocate more resources to activities such as marketing, external communication, and investor relations maintenance, crowding out the resources needed for substantive innovation and distorting the allocation of innovation resources. Second, research and development (R&D) activities tilt toward directions that can support AI narratives, reinforce intelligent labels, and align with external communication, causing the deviation of R&D direction, the weakening of process management, and a short-term orientation in outcome evaluation, thereby reducing the conversion efficiency of innovation investment. Further analysis shows that state-owned enterprises, large enterprises, and highly reputable enterprises are more likely to attract external attention and face higher expectations for technological innovation, making the inhibitory effect of AI-washing on corporate innovation more significant. This paper enriches research on the economic consequences of AI-washing and provides empirical evidence for improving the regulation of AI-related information disclosure and curbing the hype of technological concepts.
下载:
