Five major U.S. tech companies are projected to spend roughly $750 billion this year on data centers, chips, and AI infrastructure—nearly double the $400 billion they spent last year. The spending spree reveals a stark reality: artificial intelligence development isn't spreading wealth and opportunity globally. Instead, it's concentrating computational power, energy consumption, and economic control in the hands of a handful of American corporations while leaving much of the world scrambling to keep pace.
The U.S. government awarded GlobalFoundries $300 million 1 day ago to develop silicon photonics technology for more efficient AI data centers, signaling Washington's recognition that the nation's dominance in AI infrastructure requires continued public investment. Yet even as the government props up the sector, the underlying economics expose troubling inequalities. Amazon, Google, Microsoft, Meta, and Oracle now control roughly 80 percent of global computing power that drives AI, according to data cited by The New York Times. That concentration of power over a transformative technology raises urgent questions about who benefits from AI's expansion and who bears its costs.
The Energy Crisis Nobody's Talking About
Data centers consumed 64 gigawatts of electricity globally last year—roughly as much as Germany uses in a year. By the end of 2030, that consumption is expected to quadruple, eclipsing the power used by all countries in South America and Africa combined. The math is brutal: at the most advanced AI data centers, every gigawatt of power equates to roughly $40 billion to $60 billion in costs, including servers, land, connectivity, and utility hookups.
This isn't an abstract problem. Countries with limited electricity infrastructure and capital will be priced out of the AI economy. France, Germany, and other European nations are trying to encourage data center construction across the European Union, which currently controls just 5 percent of global AI computing power. But Europe has been hampered by electricity and land access constraints, permitting delays, and financing challenges. The gap between the haves and have-nots isn't shrinking—it's widening.
The Debt Trap and Market Instability
Hyperscalers are borrowing heavily to fund AI expansion, and yields are rising as investor demand cools, according to Reuters. The debt binge signals that the market may be reaching its limits. When companies need to borrow at higher costs to fund infrastructure that might not generate returns for years, it's a warning sign. Yet there's no regulatory framework ensuring these massive infrastructure investments serve the public interest or that the companies building them remain accountable to anything beyond quarterly earnings.
The chip race itself reveals another layer of concentration. The world had roughly 2.4 million "H100 equivalent" AI chips in the second year ago. That figure is expected to double roughly every nine months, putting the world on pace to have about 200 million chips by the end of 2028. Chinese companies had roughly 1.16 million H100-equivalent chips at the end of 2025, up from roughly 244,000 at the beginning of 2024, though those figures exclude smuggled chips and other offshore computing resources used by Chinese firms. The U.S.-China competition for AI dominance is real, but it's also leaving the rest of the world behind.
Who Gets Left Out
China's National Energy Administration estimated the country's electricity use for data centers will reach around 91 gigawatts in four years, or about 6 percent of total use, up from 19 gigawatts last year. Huawei, ByteDance, and Alibaba are expected to spend $111 billion on data centers and other AI investments this year. That's a massive commitment, but it's still dwarfed by U.S. spending.
Meanwhile, the Persian Gulf has pledged billions to build data centers, but the war in Iran has affected plans. Europe struggles with permitting and financing. Developing nations with lower electricity costs might seem positioned to benefit, but they lack the capital, expertise, and market access to compete. By 2029, AI infrastructure investment is forecast to top $1 trillion globally, up from $318 billion last year, according to IDC. The question isn't whether AI will transform the economy. It's whether that transformation will be shaped by democratic oversight and public interest, or left entirely to corporate boards and venture capital.
Why This Matters:
The concentration of AI infrastructure in the hands of five U.S. corporations and two Chinese tech ecosystems has profound implications for global inequality and democratic governance. When the majority of the world's AI computing power is controlled by private companies accountable primarily to shareholders, decisions about how AI develops—what it's trained on, who profits from it, what safeguards it includes—are made behind closed corporate doors. The energy demands of this infrastructure will strain grids in developing nations and wealthy countries alike, yet those nations have little say in how the technology is built. Public investment like the GlobalFoundries award shows government can shape AI development, but without stronger regulation, public oversight, and international frameworks ensuring equitable access, AI risks becoming another technology that concentrates wealth and power rather than distributing opportunity.