Artificial intelligence and global carbon inequality: Addressing the challenges and opportunities for SDG 10, SDG 12, and SDG 13

• AI increases both inter- and intra-country carbon inequality. • Trade openness helps mitigate AI’s impact on carbon inequality between countries. • National governance reduces AI’s exacerbation of intra-country carbon inequality. • AI’s impact on carbon inequality is non-linear with thresholds. • AI regulation is necessary for promoting equitable global carbon outcomes. This study examines the impact of artificial intelligence (AI) on carbon inequality (CI) in 67 countries from 1995 to 2019. The results suggest that (i) AI significantly amplifies CI both between and within countries due to its energy requirements and uneven deployment; (ii) trade openness and global value chain (GVC) positioning mitigate AI’s effect on inter-country CI, while robust governance—marked by larger government size and institutional transparency—curtails intra-country disparities; (iii) specific thresholds (trade openness > 4.74, GVC position > − 1.07, government size > 2.90, transparency > − 0.22) shift the impact of AI from exacerbating to reducing CI. The adverse effects of AI can be reversed through enhanced trade, GVC integration, and strong governance. Key policy implications: Policymakers must prioritize exceeding these thresholds to leverage AI for sustainable and equitable outcomes. This requires (a) promoting trade liberalization to spread the benefits of AI globally, reducing inter-country CI; (b) strengthening GVC participation to offset the carbon-intensive use of AI; (c) building government capacity and transparency to ensure fair adoption of AI domestically; and (d) embedding these strategies in climate policies to align AI with the long-term goals of environmental justice and the SDGs, particularly climate action (SDG 13) and reducing inequalities (SDG 10).