The booming demand for metals driven by the rapid construction of North American AI data centers, fueled by the surging need for AI computing power, is propelling steel manufacturer BlueScope's profits to skyrocket. This latest development highlights the unprecedented surge in demand for key industrial metals and minerals essential for building AI data centers, as the AI wave accelerates. As the power density of AI GPU and ASIC computing clusters increases, the expansion of AI computing power simultaneously drives demand for physical infrastructure like building structures, power transmission and distribution, and thermal management systems, boosting demand for metals such as copper, aluminum, and steel, rather than just amplifying chip demand. A key factor behind the massive profit surge for BlueScope Steel Ltd. is record growth in its North American industrial metals business, driven by the data center construction boom. AI capital expenditure is rapidly spreading from the 'silicon-based computing layer' of GPUs and HBM to the physical infrastructure layer of steel, copper, aluminum, electricity, and engineering equipment. The Melbourne-based steelmaker reported on Monday that its underlying net profit for the fiscal year ending June 30 more than doubled to A$851.2 million (US$602.4 million). Revenue in North America grew 9%, offsetting a 4% decline in Asia. From an engineering perspective, a large AI data center is first and foremost a highly industrialized 'computing factory': steel is used in the building structure, prefabricated steel components, equipment supports, racks, and peripheral facilities; copper, a vital industrial metal, is used in high/low voltage power distribution, cables, busbars, grounding, transformer connections, and liquid-cooled cold plates and heat exchangers. Aluminum, a lightweight industrial metal, has long been considered a substitute for copper, especially in cost- and weight-sensitive applications with relatively lax conductivity requirements. This substitution trend has been ongoing for over a decade and has accelerated in recent years due to resource security concerns and the push for new energy. However, in high-reliability, high-power applications, copper remains irreplaceable. Industrial metals like steel, copper, aluminum, nickel, and tin are becoming the 'core AI investment theme' beyond the AI infrastructure chain covering AI GPUs/ASICs, data center CPUs, HBM, NAND, HDD storage, 2.5D/3D advanced packaging, liquid cooling systems, optical interconnects, and data center power supply chains. Building AI applications and data centers not only requires purchasing models, chips, and high-performance AI cloud computing servers but also significant investment in the underlying energy, metals, chemicals, and resource security premiums that support the expansion of AI computing infrastructure capacity.

AI Data Center Boom 'Devours Steel'! BlueScope's Profit Doubles, North America Becomes New Growth Engine

BlueScope CEO Tania Archibald told media that North America remains the company's primary growth engine. 'We are benefiting from data center demand, and we can see its strong performance,' she said. Archibald added that the company is highly focused on organic growth and plans to consistently enhance shareholder returns over the long term. However, she noted that as the company works to control energy costs, BlueScope is also 'highly vigilant about the potential impact data centers could have on the broader energy supply.' In February, BlueScope rejected a takeover offer from Steel Dynamics Inc. and SGH Ltd., which valued the company's equity at around A$15 billion. BlueScope stated at the time that the offer undervalued the company. Since then, its share price has risen by about one-fifth. BlueScope shares fluctuated between gains and losses on Monday, trading down 0.8% as of 2:25 p.m. Sydney time. Archibald said, 'We have had no contact with that consortium for a considerable period. We very clearly rejected their last proposal because it in no way represented the fair value our shareholders deserve.' The company's North Star steel mill in Ohio also benefits from the 50% tariff on imported steel imposed by the Trump administration, aimed at addressing overcapacity from Asia. Archibald stated during an investor conference call on Monday that BlueScope expects 'continued strength in North America, a robust demand environment in Australia, and early signs of recovery in New Zealand.' She added, 'Overcapacity in some Asian markets continues to pressure regional steel spreads.' BlueScope's FY2026 presentation to shareholders explicitly lists 'Data centre and AI infrastructure build-out lifting demand' as a key support for North American demand. The North American segment's full-year underlying EBIT reached A$1.034 billion, a 101% increase year-over-year. The North Star mill's EBIT was A$805 million, maintaining 100% capacity utilization. However, profit growth was also driven by stronger steel spreads, capacity utilization, and macroeconomic growth in North America, so the profit doubling cannot be 100% attributed to AI.

BlueScope's Profit Surge Reveals the Resource Layer of Computing Infrastructure: Energy and Industrial Metals Become the Underlying Hard Assets of the 'AI Computing Arms Race'

From an AI engineering perspective, AI data centers are not an abstract 'cloud' but a highly physical capital expenditure system. GPU/ASIC computing clusters require near-endless, efficient power supply. HBM/SSD capacity expansion demands vast semiconductor materials and chemicals. The underlying construction of AI data center server rooms requires copper, aluminum, steel, as well as gas turbines, transformers, energy storage, cooling systems, and grid systems powered by natural gas. A team led by Bank of America strategist Michael Hartnett, known as 'Wall Street's most accurate strategist,' recently released a report stating that investors will continue to flock to commodity markets in the coming years. Even if the latest Middle East war ends, the global commodity market rally is expected to persist for several years into 2030. According to the BofA strategists, commodities like industrial metals represent the most logical and highest-level 'post-war' core trading theme. They bet that commodities will replace stocks as the biggest winners in the coming years, driven by investors' urgent need to hedge against risk, inflation, and a weaker US dollar. Geopolitics and the global AI race are essentially intensifying the competition for energy, rare earths, minerals, and key commodity resources. Hartnett summarized the core logic: whoever controls chips, rare earths, metals, minerals, and efficient energy wins the global AI war. This means, in BofA's view, the core pricing driver in the post-war world is no longer just interest rates and earnings, but the security of commodity supply systems, supply chain control, and fiscal expenditure expansion. The dramatic improvement in BlueScope's profits truly reveals that the AI data center construction process has moved from 'massively buying GPU/ASIC cabinets' to a new phase of 'massively accelerating the construction of industrial-grade infrastructure.' This highlights that crucial industrial metals are rapidly forming an 'AI computing infrastructure resource layer.' In other words, the end of AI is not just electricity, but 'electricity + metals + engineering capacity.' When computing resources evolve from silicon-based systems to GW-level infrastructure, industrial metals begin to acquire a structural AI attribute from their traditional cyclical identity. Among them, steel, copper, and aluminum are the most direct beta plays on AI data center physical capital expenditure, while lithium, nickel, and tin represent a second layer with varying degrees of correlation. Copper is the most typical 'AI electrification metal,' covering high-speed copper interconnects, power distribution systems, external grid expansion, and liquid cooling within data centers. Aluminum is widely used in power transmission and distribution, busbars, cables, and some cooling and structural systems. Steel corresponds to data center buildings, substations, power generation facilities, racks, and large non-residential construction. Tin's logic is more related to solder, PCBs, and electronics manufacturing, with absolute usage far smaller than steel, copper, or aluminum. Lithium and nickel benefit mainly through UPS, battery energy storage systems (BESS), and the energy storage chain supporting new power for data centers. Therefore, their 'AI purity' is significantly lower than copper's, and they are more strongly influenced by EV demand, battery chemistry systems, and mine supply cycles. Goldman Sachs, a Wall Street financial giant, estimates that global data center electricity demand will surge by 220% by 2030 compared to 2023, equivalent to adding a country among the world's top ten electricity consumers. The International Energy Agency (IEA) projects that global data center electricity consumption will rise from approximately 415 TWh in 2024 to about 945 TWh by 2030, an average annual growth of about 15%, accounting for nearly 3% of global electricity use. AI accelerator server power consumption is expected to grow by about 30% annually. US data center electricity consumption is set to increase by about 240 TWh (around 130%) from 2024 levels, contributing nearly half of the new US electricity demand by 2030. The IEA expects global data center power consumption to roughly double by 2030, with AI training/inference-specific data centers growing even faster. Meanwhile, grid bottlenecks may delay the grid connection of about 20% of planned data center capacity by 2030. These factors collectively indicate that the scarcest resource in the next phase of AI thematic investment is spilling over from 'chips' to electricity, conductors, structural materials, and energy infrastructure.