Big Tech companies are accelerating artificial intelligence development with record-breaking capital expenditure. They plan to spend over 700 billion dollars in 2026, roughly 590 billion euros, marking a 75 percent increase from 2025. Several firms released earnings reports and outlined their 2026 spending plans. Analysts confirmed these projections and highlighted their strategic importance.
Wall Street focuses on the combined capital expenditure figure. More than 700 billion dollars will flow into AI infrastructure this year. That amount exceeds Sweden’s nominal GDP for 2025, according to IMF estimates.
Global chip sales are expected to hit one trillion dollars for the first time this year. The US Semiconductor Industry Association published this forecast. JPMorgan Chase and McKinsey predict total AI capital expenditure will surpass five trillion dollars by 2030. They cite extraordinary demand for computing power as the main driver.
Capital expenditure describes spending on long-term assets like property, equipment, and technology. These investments aim to expand capacity and efficiency over multiple years. Companies capitalise costs on their balance sheets and depreciate them over time. Analysts view CapEx as a key indicator of future growth and operational strength.
The current surge confirms a trend that began in 2025. Big Tech spent roughly 400 billion dollars on AI capital expenditure that year. Nvidia CEO Jensen Huang described the expansion as the largest infrastructure build-out in human history. He repeated this statement at the World Economic Forum in Davos last month.
Hyperscalers Drive the AI Arms Race
Amazon leads the 2026 spending race with a planned 200 billion dollars in investment. That single figure exceeds the combined nominal GDP of the three Baltic states in 2025, according to IMF projections.
Alphabet follows with 185 billion dollars, while Microsoft and Meta plan 145 billion and 135 billion dollars respectively. Oracle raised its 2026 CapEx to 50 billion dollars, nearly 15 billion above earlier estimates.
Tesla plans to spend nearly 20 billion dollars to scale its robotaxi fleet and develop the Optimus humanoid robot. Elon Musk’s xAI plans to invest at least 30 billion dollars in 2026.
The company will build a 20 billion dollar data centre named MACROHARDRR in Mississippi. The state governor called it the largest private sector investment in Mississippi’s history. xAI will also expand the Colossus cluster in Tennessee, described by Musk as the world’s largest AI supercomputer.
SpaceX acquired xAI in an all-stock transaction this month. The deal valued SpaceX at one trillion dollars and xAI at 250 billion dollars. The combined entity now has a valuation of 1.25 trillion dollars, ranking among the most valuable private companies ever.
Reports indicate SpaceX plans an IPO this year. Morgan Stanley is reportedly in talks to manage the offering, which will include xAI exposure. Musk said the goal is to create an integrated innovation engine combining AI, rockets, and satellite internet. He also outlined long-term plans for solar-powered space-based data centres.
Apple remains behind in spending, with a projected 13 billion dollars. The company announced a multi-year partnership with Google to integrate Gemini models into Apple Intelligence. The collaboration will overhaul Siri and enhance on-device AI features. Apple relies heavily on this partnership instead of boosting internal AI investment.
Nvidia will report earnings and projections on 25 February. The company sells AI chips and will likely capture a large share of Big Tech spending. Data centre construction drives much of this demand. Jensen Huang previously estimated each gigawatt of data centre capacity costs 50 to 60 billion dollars, with 35 billion going to Nvidia hardware.
Investors Question the Sustainability of AI Spending
Investors show mixed reactions to the massive spending plans. Many recognise the urgency of securing an AI advantage. Others worry about the spending scale and reliance on debt financing.
Morgan Stanley estimates hyperscalers will borrow around 400 billion dollars in 2026. That more than doubles the 165 billion dollars borrowed in 2025. Analysts warn this could push high-grade US corporate bond issuance to a record 2.25 trillion dollars.
Projected AI revenue for 2026 remains far below spending. Analysts highlight risks such as rapid hardware depreciation and high energy costs. The strategy depends on future success and sustained demand. Google CEO Sundar Pichai acknowledged elements of irrationality in the spending pace.
Rothschild & Co analyst Alex Haissl downgraded Amazon and Microsoft in November. He said investors value their CapEx plans as if legacy cloud economics still applied. He warned AI economics will likely be far more expensive than expected.
Michael Burry also raised concerns, calling the AI boom a potential bubble. He highlighted unsustainable CapEx and excessive optimism. Big Tech’s AI race relies on heavy leverage, leaving outcomes uncertain. Nvidia currently benefits most, while Apple focuses on partnership-driven development instead of massive internal spending.
Europe Struggles to Keep Up
The spending surge raises urgent questions about Europe’s ability to compete. The AI race has become a battle of balance sheets. American firms mobilise nearly 600 billion euros in a single year. EU efforts do not match even the smallest US tech giant’s spending.
Brussels launched the AI Factories initiative and the AI Continent Action Plan to mobilise public and private investments. These programmes aim to strengthen Europe’s AI ecosystem and infrastructure. However, the financial gap remains stark. European sovereign cloud data infrastructure spending is forecast at only 10.6 billion euros in 2026.
That represents an 83 percent increase year-on-year but remains small compared with US investments. Mistral AI CEO Arthur Mensch said US companies build the equivalent of a new Apollo programme every year. He added Europe cannot regulate its way to computing supremacy.
Mistral represents one of Europe’s few challengers. The company aggressively expands its physical footprint. It raised 1.7 billion euros in a Series C round in September 2025, valuing the company near 12 billion euros. ASML led the round with a 1.3 billion euro investment.
Mensch confirmed a one billion euro CapEx plan for 2026. Mistral also announced a 1.2 billion euro investment to build a data centre in Borlänge, Sweden. The project partners with EcoDataCenter and aims to deliver sovereign compute under strict EU data standards. Sweden’s green energy will power the facility.
The data centre will open in 2027 and provide high-performance computing for next-generation AI models. It is Mistral’s first project outside France and a core pillar of European data sovereignty.
US tech giants also offer “sovereign-light” solutions in Europe. They launch localised cloud zones in Germany and Portugal, promising data residency. Critics argue these remain dependent on US parent companies, leaving Europe vulnerable to American economic and policy shifts.
As 2026 unfolds, the stakes are clear. The United States bets heavily on AI dominance and its credit capacity. Europe relies on targeted investments and regulation to carve out a sovereign niche in a world increasingly shaped by American technology.
