The Effect of AI Adoption on Earnings Quality: Evidence from S&P 500 Technology Firms
0 viewsAbstract
This study examines whether the extent to which firms discuss Artificial Intelligence (AI) in their annual reports is associated with the quality of their reported earnings. Traditional financial reporting depends heavily on managerial judgment, while previous research has shown that the language used in corporate filings can contain information beyond the financial statements themselves (Loughran & McDonald, 2011). However, relatively little empirical research has directly examined whether AI-related disclosure is connected to earnings quality, particularly during the period following the rapid rise of generative AI. To investigate this relationship, I constructed a balanced panel of S&P 500 technology firms covering the study period. AI disclosure intensity was measured through keyword-frequency analysis of Form 10-K filings, while earnings quality was proxied by the absolute value of discretionary accruals estimated using the Modified Jones Model (Dechow et al., 1995). The analysis controls for firm size, leverage, profitability, growth, and year effects. The results indicate a statistically significant negative association between AI disclosure intensity and absolute discretionary accruals. Because lower absolute discretionary accruals are interpreted as higher earnings quality, the findings suggest that firms with more extensive AI-related disclosure tend to report higher-quality earnings. These results contribute to the growing literature on textual analysis of corporate disclosures and provide potentially useful information for investors, auditors, and regulators evaluating AI-related corporate communication.
Keywords
Artificial Intelligence, Discretionary Accruals, Earnings Quality, Financial Reporting, Textual Analysis