The world’s largest technology companies have invested more than $1.1 trillion in capital expenditure since the emergence of the artificial intelligence boom in 2023, reflecting an extraordinary transformation in how the sector is allocating resources as competition to dominate AI accelerates.
According to the latest earnings reports released over the past two weeks, Google, Amazon, Microsoft and Meta collectively spent $1.1 trillion on capital investments between the start of 2023 and the end of June. The spending marks a significant shift for companies that historically operated as relatively asset-light businesses but are now committing vast sums to physical infrastructure, including data centres, advanced semiconductor chips and the energy required to power them.
The investment surge is set to continue. The four companies are projected to spend a combined $745 billion on capital expenditure during this year alone after both Google and Amazon raised their investment forecasts in the latest quarter. Analysts say there is little indication that the pace of spending will slow as the companies seek to strengthen their positions in the rapidly evolving AI market.
“There is basically no end in sight for the growth in capex,” said Rishi Jaluria, an analyst at RBC Capital. He noted that investors expect the companies to maintain a careful balance between investing aggressively in AI while preserving the strengths of their established businesses.
The scale of the investment also depends on the continued expansion of AI developers OpenAI and Anthropic, both of which have entered into substantial long-term commitments to purchase computing capacity as they pursue future public listings. At the same time, the rapid build-out of AI infrastructure has placed increasing strain on global supply chains, contributing to shortages of memory chips and higher component costs.
Those supply constraints have affected companies outside the leading AI investment race. Apple warned that rising costs would reduce sales and profit margins, contributing to a 6.3 per cent decline in its share price on Thursday.
Despite the enormous expenditure, recent financial results suggest that AI investments are beginning to generate stronger revenue growth, particularly in cloud computing. Google, Amazon and Microsoft all reported accelerating growth in their cloud businesses as demand for AI computing services increased among both AI developers and corporate customers adopting the technology.
Meta, which does not operate a cloud computing business, said artificial intelligence had improved the effectiveness of its advertising platform, helping lift quarterly revenue by 28 per cent year on year to $61 billion. Chief executive Mark Zuckerberg also revealed that the company had received numerous offers to lease its computing capacity at prices significantly above its costs, although he maintained that selling AI-powered services would remain substantially more profitable than renting out computing infrastructure.
Not all investors were convinced by Meta’s strategy. The company’s shares fell 8 per cent following its earnings announcement, with some analysts citing the absence of a clearly defined plan for commercialising its expanding AI infrastructure. Dec Mullarkey, managing director at SLC Management, said investors increasingly wanted to see evidence that heavy AI spending was translating into measurable financial performance rather than simply supporting growth ambitions.
Google also faced investor scepticism despite adding $11 billion in annual revenue to its cloud business. The company reported negative free cash flow of $6 billion during the quarter, its first period of negative free cash flow since becoming a public company more than two decades ago, prompting a sell-off in its shares.
The technology groups also disclosed a sharp increase in long-term financial commitments linked to AI infrastructure. Google expanded its future commitments by approximately $500 billion over three months, largely through long-term agreements covering technical infrastructure and energy supplies for data centres. Meta added $233 billion in new commitments during the quarter, including data centre leases, infrastructure purchases and new debt, before signing a further $68 billion in data centre leases during July. Microsoft, meanwhile, entered into more than $130 billion in new data centre lease agreements during the second quarter.
Together, Google, Microsoft and Meta assumed almost $900 billion in new AI-related obligations during the three-month period, while Amazon has yet to disclose comparable details. Executives across the sector acknowledged that sustained investment would continue to weigh on free cash flow in the near term. Combined free cash flow for the four companies fell to a decade low of just $7 billion during the period, with only Microsoft and Meta generating more cash than they spent.
Amazon chief executive Andy Jassy said the company expected free cash flow pressures to persist while simultaneously constructing numerous data centres, noting that new facilities typically require around two years before becoming operational and generating revenue. The extended period between infrastructure investment and financial returns means investors may have to wait several years before the full benefits of the industry’s record AI spending become evident, even as companies continue to deepen their commitment to the technology.
