Investor Attribution Under Algorithmic Uncertainty: How Corporate Social (Ir)Responsibility Shapes Market Reactions to AI Adoption

Artificial Intelligence (AI) adoption is reshaping organizational decision-making. However, market reactions to such adoption vary widely. This study examines how prior corporate social performance (CSP)—specifically corporate social responsibility (CSR) and corporate social irresponsibility (CSIR)—influences investor responses to AI adoption. We also investigate whether the specificity of AI information moderates these effects. To test our hypotheses, we conducted three complementary studies. The first is a longitudinal analysis of data from Chinese publicly listed firms from 2017 to 2022. The other two are scenario experiments designed to examine investors' attributions as the underlying mechanisms. Our findings show that the market consequences of AI adoption are complex and not universally positive. Instead, its market impact depends on a firm's prior CSP. Investors are more likely to attribute AI adoption to altruistic motives when firms have strong CSR histories, thereby increasing market performance. By contrast, for firms with CSIR histories, investors tend to attribute AI adoption to self-serving motives, which backfires and decreases firms' market performance. We further find that the specificity of AI-related disclosure strengthens the positive impact of AI adoption for firms with strong CSR records, while acting as a buffer to mitigate the negative impact for firms with CSIR records. These findings are robust across multiple empirical specifications, including instrumental-variable estimation, selection-bias corrections, and alternative measures. Overall, the findings underscore the strategic importance of treating AI adoption as a sociotechnical governance challenge, aligning implementation with a firm's social reputation, and communicating these initiatives transparently to foster favorable market perceptions.