Data Security Risk Exposure in Chinese Enterprises: A Large Language Model Approach

With the rapid expansion of the digital economy, data security risk has become an increasingly important economic concern. Using large language models to analyze more than 560,000 analyst reports from 2006 to 2022, we develop firm-level measures of data security risk exposure for Chinese listed companies. Extensive empirical tests show that the measures capture systematic variation in exposure across time, regions, and industries and significantly predict firms’ subsequent vulnerability status. Firms with higher exposure tend to be larger, more profitable, and more liquid and to have greater growth opportunities. This study provides a forward-looking quantitative tool for enterprises and regulators to identify and mitigate potential risks during the digital transformation process.