DOI: 10.11648/j.stpp.20261002.13 ISSN: 2640-4621

Geopolitical Sustainability of Powering Distributed Artificial Intelligence in Pakistan

Babar Chaudary, Petri Pulli
Show PDF Cite
Artificial intelligence (AI) is increasingly deployed across a continuum ranging from stationary data centers and semi-stationary computing platforms to autonomous vehicles, robots, wearable devices, implants, and Internet of Things nodes. Across this distributed continuum, reliable power supply and the availability of critical raw materials become system-level constraints. This study investigates how power-supply requirements of distributed AI propagate into material and geopolitical dependencies, with particular attention to Pakistan. A Design Science Research approach is applied to a representative wearable-edge-cloud healthcare architecture for assisting visually impaired persons. The artefact is decomposed into power, sensing, communication, computation, and interaction subsystems; representative components are mapped to critical materials; materials are grouped by affinity; and mining, refining, and geographic concentration are examined using peer-reviewed literature and authoritative mineral and energy sources. The analysis identifies recurring dependencies on lithium, graphite, copper, cobalt, rare-earth elements, semiconductor materials and, for prospective hydrogen configurations, platinum-group metals. It also shows that cloud offloading shifts rather than removes material dependency because communication, data center, grid, and backup-energy infrastructures introduce additional upstream requirements. For Pakistan, the study distinguishes documented mineral production and resource development from prospective geological occurrences and downstream refining capability. The results indicate that geological endowment alone does not ensure access to technology-grade materials. The paper therefore contributes a traceable analytical chain linking AI service architecture, power subsystems, component/material dependencies, geographic supply concentration, and geopolitical vulnerability. Scenario analysis is used as a stress test rather than a prediction, highlighting the value of diversified sourcing, processing capability, strategic cooperation, and adaptable system design for resilient distributed AI deployment.

More from our Archive