**Introduction** Artificial intelligence (AI) places growing demands on physical resources because advanced AI systems require large-scale computing infrastructure, continuous electricity supply, semiconductor manufacturing, cooling systems, and access to critical minerals. As AI models become larger and more widely deployed, the physical infrastructure supporting them expands rapidly, increasing pressure on energy systems, industrial supply chains, and natural resources. **Contextual background** AI systems depend on data centers equipped with advanced processors capable of performing enormous volumes of calculations. Training and operating these models require constant electricity, extensive cooling, and sophisticated semiconductor technologies. These demands make AI development more closely tied to industrial production systems and resource-intensive supply chains than earlier digital technologies. As AI capabilities become central to technological leadership and geopolitical competition, governments are treating advanced computing infrastructure and semiconductor capacity as strategic assets[1]. **Reasons why AI increases demand for physical resources** **1.** **AI significantly increases electricity consumption** AI systems consume large amounts of electricity because data centers must operate continuously to process and store massive volumes of data. Global electricity consumption from data centers is projected to rise significantly by 2030, driven largely by the expansion of AI workloads and high-performance computing[2]. This growing electricity demand places pressure on power grids, generation capacity, and energy infrastructure. Governments and firms increasingly invest in renewable energy, transmission systems, and backup power generation to support AI-related infrastructure growth. **2.** **AI depends on advanced semiconductor manufacturing** AI systems rely heavily on advanced semiconductors such as graphics processing units (GPUs) and AI accelerators. Producing these chips requires highly resource-intensive manufacturing processes involving ultrapure water, specialized chemicals, rare gases, and precision industrial equipment. The concentration of semiconductor production in a limited number of economies has increased concerns over supply-chain security and technological dependence. Governments are increasingly expanding industrial policies, subsidies, export controls, and strategic investment to strengthen domestic semiconductor manufacturing capacity and reduce vulnerability to external supply disruptions[3]. **3.** **AI increases demand for critical minerals** The expansion of AI infrastructure raises demand for critical minerals used in semiconductors, electricity systems, batteries, and cooling technologies. Minerals such as copper, lithium, nickel, cobalt, and rare earth elements are essential for both data center infrastructure and the broader electrification systems that support AI growth[4]. As demand rises, governments increasingly view critical mineral supply chains as strategic assets linked to economic security and industrial competitiveness. This has accelerated efforts to diversify mineral sourcing, strengthen processing capacity, and secure long-term supply agreements. **4.** **AI infrastructure requires substantial water and land resources** Large AI data centers generate significant heat and therefore require extensive cooling systems, many of which depend on large volumes of water. In regions with concentrated data center development, this can increase pressure on local water supplies. AI infrastructure also requires large industrial sites with reliable electricity access, high-speed connectivity, and proximity to transmission networks. As hyperscale data centers expand, land use and infrastructure planning become increasingly important components of AI development strategies. **Conclusion** AI places growing demands on physical resources because its development depends on energy-intensive computing infrastructure, advanced semiconductor production, critical minerals, and large-scale industrial facilities. As AI adoption accelerates globally, these resource requirements are increasingly shaping energy policy, industrial strategy, supply-chain security, and infrastructure investment. AI is therefore becoming not only a digital technology issue, but also a major industrial and resource management challenge.
