AI Arms Race Burning Too Much Cash? Tech Giants' Cash Flow Warnings Are Sounding, and One Major Player Is Making Markets Nervous

The AI frenzy continues to push up investment levels among U.S. tech giants. While Microsoft, Alphabet, Amazon, Meta, and Oracle—the so-called 'hyperscale cloud providers'—are steadily showcasing results from their AI investments, massive capital expenditures are rapidly eroding free cash flow. Markets are now questioning whether the profit growth from AI can justify this unprecedented spending race.

According to Reuters’ analysis of LSEG market forecasts, if current trends continue, the combined capital expenditures of these five tech giants will exceed free cash flow for the first time by 2027—meaning AI infrastructure investments will consume all cash that companies could otherwise freely use.

For every $1 increase in cash flow, $1.57 is invested in capital expenditure

Data shows that by 2027, the five companies’ operating cash flow will increase by approximately $340 billion compared to 2025, while capital expenditures are projected to rise by about $534 billion. In other words, for every $1 increase in operating cash flow, companies must invest an average of $1.57 in new capital.

Markets will closely watch upcoming tech earnings reports. After Alphabet releases its results this week, investors will monitor whether cloud and AI-related revenues can sustain high growth and keep pace with rising expenses.

Stock performance already reflects these concerns. While large tech stocks were the biggest beneficiaries of the recent AI boom, over the past year, all except Alphabet have underperformed the S&P 500 index.

AI is transforming tech giants’ business models—Is the 'light-asset' era ending?

Shay Boloor, Chief Market Strategist at Futurum Equities, says the market is still underestimating the structural impact of AI on large tech companies.

He notes that in the past, big tech firms enjoyed high valuations as 'light-asset' businesses because revenue growth far outpaced capital investment. But the AI era is different: software, advertising, and cloud services now depend on massive data centers, servers, and networking equipment. Business models are gradually shifting toward a new hybrid model that balances software and hardware.

Since companies do not disclose detailed AI investment figures, LSEG’s statistics cover overall capital expenditures. However, multiple company executives have stated that current spending on data centers, servers, and network infrastructure is largely driven by AI demand.

Notably, market estimates for investment levels continue to be revised upward. Analysts initially forecasted around $485 billion in annual capex for the five firms at the start of the year, but that figure has now risen to approximately $730 billion—indicating the AI arms race is still rapidly intensifying.

AI revenues are materializing—but cash flow pressure is mounting

Some companies have already demonstrated that AI is generating real revenue. For example, Microsoft’s AI-related business now has an annualized revenue run rate exceeding $37 billion. Amazon’s AWS cloud division saw a 28% year-over-year revenue increase in Q1, reflecting continued AI-driven growth in cloud services.

However, investors care more about whether AI can generate sustainable cash flow, not just revenue.

Microsoft’s latest quarterly operating cash flow was $35.8 billion, but capital expenditures—including financing leases—reached $37.5 billion, meaning spending exceeded cash inflow during the quarter.

David Russell, Global Market Strategist at TradeStation, says a company’s purpose is to generate profit, not to continuously invest capital. If capital spending keeps consuming cash, even growing profits may not justify such massive investments.

Amazon shows a similar trend: although its trailing 12-month operating cash flow rose 30% year-on-year to $148.5 billion as of Q1, free cash flow dropped to just $1.2 billion, showing that heavy investments are significantly squeezing available funds.

Oracle becomes a market warning signal

Among the five, market concerns are most pronounced for Oracle. Its free cash flow has turned negative, and its stock has fallen 36% year-to-date. The company announced plans to raise $45–50 billion through debt and stock issuance to continue expanding its cloud infrastructure.

LSEG data shows Oracle’s capital expenditure as a percentage of operating cash flow has surged from 47% in fiscal 2022 to 174% in fiscal 2026 (ending May this year). In the latest fiscal year, capital expenditures reached $55.7 billion, far exceeding its $32 billion in operating cash flow—indicating investment levels have surpassed the company’s own cash-generating capacity.

Has AI investment gone too far?

In contrast, Microsoft, Alphabet, and Meta still maintain strong free cash flow, sufficient to support dividends and share buybacks, with no immediate impact on shareholder returns. However, some market participants worry that if AI investments remain high and commercialization lags, future buyback programs could be scaled back.

Freddy Lavric, Senior Trader at Winthrop Capital Management, says the next 2–3 years will be a critical observation period. Tech giants must prove that AI truly drives new revenue, improves profit margins, and enhances cash flow. If these financial benefits don’t become clearly evident, markets will begin re-evaluating whether AI investments have gone too far.