Abstract: Corporate AI washing, a decoupling between AI-related talk and action, has become an important issue threatening innovation ecosystems, yet its performance-feedback drivers remain underexplored. Integrating performance feedback theory and impression management theory, this study examines how dual performance aspiration gaps affect corporate AI washing and identifies their boundary conditions. Using a sample of Chinese A-share listed firms from 2011 to 2023, we test the hypotheses with two-way fixed-effects regressions. The results show that historical aspiration surplus and industry aspiration shortfall significantly increase corporate AI washing, whereas historical aspiration shortfall and industry aspiration surplus significantly inhibit it. When board independence or regional AI attention is high, firms that meet industry aspirations show a weaker positive effect of historical aspiration surplus on AI washing and a stronger negative effect of historical aspiration shortfall. Heterogeneity analysis further shows that these relationships are particularly pronounced among high-tech enterprises and non-digital enterprises. Based on these findings, governments should establish a dual institutional-cognitive regulatory system to supervise and standardize corporate AI-related technology disclosures. Stakeholders should strengthen their ability to identify AI washing and pay particular attention to firms with historical aspiration surpluses and industry aspiration shortfalls. Firms, in turn, should optimize executive incentive mechanisms and strengthen the internal monitoring role of independent directors. A tripartite governance mechanism involving governments, stakeholders, and firms can effectively curb corporate AI washing.

Key words: artificial intelligence, AI washing, dual performance aspiration gaps, board independence, regional AI attention