Artificial Intelligence and Climate Risk Shocks.

Abstract

Artificial intelligence (AI) has been widely applied across various fields and has demonstrated effectiveness to some extent. However, some scholars have raised concerns about its ethical implications and potential rebound effects, particularly in the context of climate issues. To address these debates, we obtained cross-national panel data for 51 countries from 1996 to 2023 through the ISETS Energy Finance Network, WIPO, and World Bank databases. A two-way fixed effects model was used to examine the relationship between AI and physical climate risks. The findings are as follows: (1) AI can effectively govern physical climate risks. (2) The governance effect of AI on physical climate risks comes from risk sensing, risk seizing, and action integration. (3) Compared to AI technologies and application fields, AI functional applications have a greater governance impact. (4) Political and economic integration alliances show a stronger AI governance effect compared to purely economic alliances. (5) The governance effect of AI intensifies as physical climate risks increase. These results provide theoretical support and empirical evidence for government efforts to promote AI applications and achieve climate governance. These findings provide theoretical support and empirical evidence for government initiatives to promote AI applications in climate governance.