Spending by Big Tech on artificial intelligence has soared to unprecedented levels.
In recent earnings calls, major technology companies have significantly increased their forecasts for capital expenditures. Alphabet, the parent company of Google, announced on Wednesday its intention to double its capex by 2026, reaching nearly $185 billion. On Thursday, Meta revealed plans for a colossal $200 billion in capital spending, far exceeding Wall Street expectations. Last week, Amazon projected its full-year capex could climb to $135 billion. Combined with Microsoft’s rising estimates, the total spending from these firms exceeds a staggering $630 billion.
This investment represents a highly concentrated bet by Big Tech. Not only is the dollar amount vastly higher, but the focus is singularly on scaling AI computing power, rather than being spread across a portfolio of strategic initiatives.
The scale of corporate investment in AI infrastructure now matches the economic output of some of the world’s largest nations, comparable to the annual GDP of countries such as Sweden and Israel. Capital expenditure funds major infrastructure like data centers, servers, and power systems that are essential for the AI expansion. These facilities—some anticipated to be as large as a football field or even bigger—demand enormous resources and energy for construction, maintenance, and operation.
“We’ve never invested this much in anything before,” stated Gil Luria, managing director and head of technology research at D.A. Davidson. “But we’ve also never had a technology this promising before.”
Data centers in your shopping mall
As companies pour money into physical data center infrastructure, some experts predict the next wave of construction will extend into local communities. “I firmly believe that the ‘Stranger Things’ mall where they battle the creature will be converted to a data center,” said Brent Thill, an analyst at Jefferies.
The current AI infrastructure push is historically unmatched in scale. However, Luria notes the spending directly mirrors existing market demand. “It’s an unprecedented buildout,” Luria said. “But it’s really being done in conjunction with the growth in demand.” He highlighted that demand backlogs for Amazon, Meta, and Microsoft have hit record levels, with Microsoft’s order backlog swelling to $625 million, partly driven by OpenAI.
Thill explained that this buildout aims to solve the AI industry’s current bottleneck: physical infrastructure. He noted the constraint has moved from semiconductors to energy supply, and now to a shortage of the actual physical structures. “It went from a chip shortage, a GPU shortage,” Thill said. “Now, it’s a physical shell shortage.”
Market skepticism and the software squeeze
However, as firms channel vast sums into AI infrastructure, it has triggered a reassessment of software company valuations. This led to a major weeklong selloff in tech stocks and cryptocurrencies, as AI progress raises questions about the future relevance of some software technologies. While companies express confidence in AI’s potential, the technology has yet to deliver substantial profits, and investors are reacting to uncertainty about its tangible value. Combined with disappointing employment figures, investor anxiety over AI contributed to a loss of nearly $1 trillion in value from software and services stocks. Not all are worried, including Nvidia CEO Jensen Huang, who has previously dismissed calls for immediate returns on investment.
According to Luria, investors in tech giants are becoming increasingly anxious as these companies deploy virtually all their available capital to fund the infrastructure expansion. He stated that shareholders are looking for profits, not further capital commitments. “‘We understand that you want to invest all this money, but you’re investing all our money, you’re taking all your cash and all your cash flow and investing it,’” Luria said, summarizing the shareholder perspective.
Despite the market downturn, Big Tech is wagering on substantial returns from AI. “We’re in a game of leapfrog now,” Thill said. “You have three to four big public vendors that are all lined up for this prize.”
Regarding the rationale for the aggressive buildout, Thill suggested that given the current demand for AI data centers, the primary risk for tech firms is not doing enough. Any excess capacity would still find a market.
“Even if you overbuild,” Thill said, “there’s so many people that would buy that overbuild even if they couldn’t sell it to their clients. Other people would want to procure it: state, local governments, [and] federal governments.”