Free cash flow sounds the alarm! The four major CSPs in the U.S. have accumulated capital expenditures exceeding $1.1 trillion. Where is the endpoint of the AI gamble?

The four major cloud service providers in the United States—Google, Amazon, Microsoft, and Meta—have collectively surpassed $1 trillion in capital expenditures since the AI boom began in 2023, demonstrating the tech giants' strong commitment to artificial intelligence (AI).

According to the latest financial reports released by the four companies in the past two weeks, their combined capital expenditures from the beginning of 2023 to the end of June 2026 have reached $1.1 trillion, primarily invested in AI data centers, advanced semiconductors, servers, and power infrastructure. Market estimates suggest that the total capital expenditures for the four companies will further rise to approximately $745 billion this year, with both Google and Amazon raising their annual investment plans this quarter, indicating that the competition for AI infrastructure continues to heat up.

RBC Capital analyst Rishi Jaluria stated that there is currently no end in sight for the growth of capital expenditures among large tech companies. The market is no longer just focused on whether companies will continue to invest in AI, but rather on how they can maintain profitability and shareholder returns while making substantial investments.

The AI arms race also heavily relies on whether AI model developers like OpenAI and Anthropic can continue to raise funds. In recent years, both companies have committed to large-scale, long-term procurement of computing resources from cloud providers. Their ability to secure funding to fulfill these contracts will impact the entire AI infrastructure investment cycle.

The four major CSPs' free cash flow turned negative this year, further expanding in Q2

On the other hand, large-scale investments continue to exert pressure on the supply chain. The demand for advanced chips, high-bandwidth memory, and other components in AI data centers has surged, leading to supply shortages and increased overall semiconductor costs.

According to Nikkei, the investment scale of the four major CSPs in AI and other areas in the second quarter exceeded the cash flow generated by their core businesses by nearly $100 billion. Additionally, investments in AI infrastructure are rapidly eroding corporate cash flow.

Reports indicate that the four major CSPs had a combined negative cash flow of $94.6 billion from April to June this year. Compared to the first quarter, which marked the first negative quarter since 2018, the cash flow gap further widened in the second quarter, with total free cash flow remaining at approximately $7 billion, a near ten-year low, with only Microsoft and Meta maintaining positive figures.

At the same time, the scale of hidden liabilities outside of financial reports is rapidly increasing. Nikkei pointed out that the combined off-balance-sheet hidden liabilities of the four major CSPs along with Oracle reached $1.65 trillion in the latest quarter, exceeding the approximately $1.35 trillion in on-balance-sheet interest-bearing liabilities. Among them, Meta's off-balance-sheet debt is particularly large, amounting to about $420 billion, nearly three times its on-balance-sheet liabilities.

Market concerns about whether AI investments can yield substantial returns

Currently, the market's attitude towards the tech giants' continued expansion of AI investments has become cautious, and the frequent issuance of corporate bonds by companies in recent years has also raised financing costs Currently, another important observation in the market is how long the cloud demand driven by AI can be sustained. Although corporate clients have begun to allocate budgets to promote AI-driven digital transformation and operational reform, it remains to be seen whether this will ultimately bring sufficient business benefits.

From the recent financial report performances, the four major companies have begun to show differences in converting AI investments into core business profitability, especially contrasting Amazon and Microsoft, which focus on cloud services, with Meta, which lacks enterprise cloud business.

The market expects that Microsoft's AI investments have started to translate into core business profitability, leading to a 16% surge in stock price the day after the financial report was released, with a single-day market value increase of approximately $448 billion, setting a record for the largest single-day market value increase in history. In contrast, Meta, which relies on social media platforms, has seen its operating profit decline for the first time in three years due to rising costs of AI talent recruitment and AI computing usage.

Dec Mullarkey, Managing Director of SLC Management, pointed out that Meta has yet to present a clear strategy for computing power leasing, making it difficult for investors to assess the return on its AI investments, which is also a significant reason for the stock price drop after the financial report was released. In contrast, Google, Microsoft, and Amazon have demonstrated actual revenue growth from AI investments through cloud services, which is more convincing for the market to continue investing capital.

• Big Tech AI spending spree tops $1tn
• Four US tech giants bled $95bn in cash in Q2 on soaring AI investments

(Source of the main image: Unsplash)

Further Reading:

• Qualcomm completes acquisition of Modular, accelerating generative AI and agent-based AI from edge to cloud layout
• Apple's financial forecast falls short of expectations, plunging 7%, relinquishing the market value throne to Nvidia