<p>This study investigates convergence patterns in natural resource efficiency across 42 countries in the Global North and South from 2000 to 2022. It further examines the influence of artificial intelligence and geopolitical risks on the formation of convergence clubs. Lastly, the study employs a regression discontinuity design to assess the causal impact of the United Nations 2030 Agenda on resource efficiency. The results reveal an uneven distribution of resource efficiency across the sample, with Global South countries exhibiting lower efficiency levels than those in the Global North. The Phillips and Sul method rejects the overall convergence, but identifies three final clubs in resource efficiency. The results of Ordered Probit regression show that nations exhibiting higher levels of AI tend to converge toward higher resource efficient countries. In contrast, countries experiencing greater geopolitical challenges tend to converge with those exhibiting lower resource efficiency levels. The Fuzzy regression discontinuity analysis shows a discontinuity in resource efficiency at the cut-off, indicating that the 2030 UN agenda has effectively enhanced the resource efficiency likelihood in the North–South countries. The findings suggest that policymakers should design targeted strategies that promote AI adoption and manage geopolitical risks to improve resource efficiency and advance the 2030 Agenda, particularly SDG 12 and SDG 9.</p>