The massive capital expenditure plans of Big Tech have become a defining story of the year, with total spending by Google, Meta, Amazon, and Microsoft expected to reach $750bn in 2025. That sum represents roughly half the annual budget of the entire UK government, underscoring the scale of the bet on artificial intelligence infrastructure.
This investment splurge, already outpacing prior budgets, is slated to grow further next year. The four firms are racing to secure competitive positions in the AI revolution, pouring funds into data centers, chips, and energy systems. However, the sheer velocity of spending raises questions about long-term asset utilization.
A central risk is depreciation. As hardware—especially specialized AI chips and data centers—ages rapidly or becomes obsolete, companies may face outsized write-downs. The gap between capital commitments and eventual revenue generation from these assets could widen, creating a financial time bomb for balance sheets.
Market analysts caution that while the upfront spending fuels growth narratives, the accounting reality will eventually catch up. Slower-than-expected AI adoption or shifts in technology could accelerate depreciation, forcing earnings revisions. The paradox: the very spending meant to secure dominance might erode future profitability.
Yet proponents argue that early movers will capture outsized returns, making depreciation a manageable cost of long-term leadership. The tension between short-term financial risk and long-term strategic imperative now defines the sector's outlook.