We document “foreign bias” in AI financial predictions, reversing the classic home bias. U.S.-based ChatGPT is systematically more optimistic than China-based DeepSeek about Chinese firms—in price predictions and directional forecasts—yet significantly less accurate. Evidence supports an information-availability mechanism: bias is strongest when U.S. media coverage of Chinese firms is limited and attenuates for cross-listed firms. Crucially, injecting Chinese news eliminates the prediction gap. Both models produce similar forecasts for U.S. firms, consistent with broader worldwide coverage. LLMs trained in different information environments can create divergent signals, with implications for investors and policymakers as AI increasingly intermediates global markets.

This paper examines how AI investment and data assets affect corporatecredit risk. Using Chinese listed firms, we construct four complementary measures ofAI investment, asset-based, labor-based, LLM-based, and text-based, and link them tofirms’ distance-to-default. We find that benchmark-level AI investment reduces defaultrisk, while excessive ffrm-speciffc investment increases it by eroding profitability andreffecting risk-taking and competitive pressure. The dominance of this adverse effectyields a negative overall relation between AI investment and credit risk. Cash flow riskis the transmission channel: benchmark-level AI improves cash ffow quality, whereasexcessive investment worsens it. High-quality data assets complement benchmark-levelAI by stabilizing cash ffow, but this benefit fades once investment becomes excessive.Overall, the impact of AI on credit risk depends on both investment intensity and dataquality, operating primarily through cash flow dynamics.