“Earnings Rise, but Cash Flow Turns Negative”: Google Faces AI Investment Backlash as Big Tech Spending Race Deepens Financial and Profitability Pressure

Google maintains aggressive investment strategy despite negative free cash flow
Big Tech AI spending intensifies amid fears that delayed investment means falling behind
Shadow liabilities continue to expand while the path to AI monetization remains uncertain

Alphabet, Google’s parent company, has recorded negative free cash flow for the first time since going public. Although the company continued to deliver strong revenue growth, led by its cloud business, capital expenditure on artificial intelligence infrastructure exceeded the cash generated from operations and significantly increased its financial burden.

As competition among major technology companies to secure AI infrastructure becomes increasingly intense, concerns over the economic effectiveness of these investments and the prospects for eventual monetization continue to weigh on the market.

Alphabet’s Cash Flow Comes Under Pressure

Alphabet announced on July 22, local time, that its revenue for the second quarter of 2026 increased by 24% year on year to $119.8 billion, exceeding the $116.9 billion consensus estimate compiled by the London Stock Exchange Group. Operating income rose by 30% to $40.77 billion, while earnings per share surged by 294% to $9.11. The cloud business was the primary driver of this growth. Google Cloud revenue increased by 82% from a year earlier to $24.8 billion, while revenue from Google Search and other related businesses, the company’s principal source of income, reached $63.3 billion. YouTube advertising revenue rose to $11.1 billion, with both segments recording growth rates in the double digits.

Alphabet’s second-quarter capital expenditure reached $44.9 billion, broadly in line with the $44.8 billion estimate compiled by StreetAccount. The company’s free cash flow, however, fell to negative $5.86 billion as the continuing scale of investment exceeded the cash generated through operating activities. Negative free cash flow means that a company is spending more on factories, equipment, and other capital assets than it earns from its ordinary business operations. Alphabet had never reported negative free cash flow since its initial public offering in 2004. Its debt burden has also increased sharply as the company repeatedly issued bonds to finance AI investment. During the second-quarter earnings call, Chief Financial Officer Anat Ashkenazi said Alphabet’s debt had risen from approximately $16 billion a year earlier to nearly $100 billion.

Despite this pressure, Google indicated that its capital expenditure would exceed its previous forecast. Ashkenazi said that although the company had expanded capacity substantially during the past three years, customer demand continued to outpace the scale of investment. Alphabet consequently raised its full-year capital expenditure outlook from an earlier range of $180 billion to $190 billion to between $195 billion and $205 billion. The company’s shares fell by more than 4% in after-hours trading following the announcement, reflecting investor concern that even strong operating growth may not be sufficient to offset the financial consequences of the infrastructure expansion.

Big Tech’s Spending War Continues

Google has adopted such an aggressive investment strategy because its largest competitors are expanding their own AI spending at a similarly rapid pace. Companies that fail to secure computing capacity in advance risk encountering difficulties across model development, service launches, and customer acquisition. If a company delays investment while competitors expand infrastructure, it may lose both supply capacity and market share simultaneously. This dynamic explains why major technology companies continue to prioritize AI infrastructure even at the expense of short-term cash flow and profitability.

Amazon has presented a capital expenditure plan closest in scale to Google’s, with expected full-year spending of approximately $200 billion. Market observers expect a substantial portion of the total to be directed toward Amazon Web Services data centers and AI computing infrastructure. Amazon has accelerated efforts to secure Nvidia graphics processing units and construct additional data centers while increasing production and deployment of its internally developed Trainium and Inferentia chips. The company aims to reduce its dependence on general-purpose GPUs, lower the cost of AI training and inference, and provide customers with more affordable computing services.

Microsoft has announced plans to spend approximately $190 billion on capital investment during the year, concentrating expenditure on Azure data centers, GPU and CPU servers, proprietary AI accelerators, and network equipment. Meta has also raised its annual capital expenditure forecast from a previous range of $115 billion to $135 billion to between $125 billion and $145 billion. The company plans to direct the funds toward large AI training clusters, data centers, GPU servers, and networking infrastructure as it expands its internal computing foundation for advanced models, personalized recommendation algorithms, smart glasses, and other next-generation products.

Concerns Grow Over the Consequences of Excessive Investment

The problem is that the risks carried by major technology companies are increasing alongside the investment boom. The Nihon Keizai Shimbun reported on July 21 that off-balance-sheet obligations accumulated by Alphabet, Microsoft, Amazon, Meta, and Oracle through special-purpose vehicles and leasing arrangements had recently reached approximately $1.65 trillion. These financing structures, widely used to build data centers and secure high-performance semiconductors, have created what is increasingly described as a large pool of “shadow debt.”

Major technology companies often allow a special-purpose vehicle to borrow money and construct data centers, servers, and other infrastructure, then lease the completed facilities over an extended period. If a contract remains at a pre-completion or pre-delivery stage, or is classified as a service arrangement rather than a lease, the associated right-of-use asset and lease liability may not immediately appear on the company’s financial statements under US accounting standard ASC 842 or international standard IFRS 16. Funding for these structures is commonly supplied by private credit providers, infrastructure funds, pension funds, and commercial banks, while project financing is frequently combined with long-term leases and power purchase agreements.

Such arrangements do not necessarily violate current accounting standards, but they can prevent financial statements from fully reflecting the rent, purchasing commitments, and other payment obligations that companies will bear over many years. As a result, the economic liabilities associated with AI infrastructure may be substantially greater than the debt formally reported on corporate balance sheets. If demand fails to develop as expected, utilization rates remain low, or the profitability of AI services is delayed, these long-term commitments could create sustained pressure on cash flow and credit quality.

Another concern is that the economic gains generated by the AI ecosystem remain heavily concentrated among semiconductor companies. Jim Covello, head of global equity research at Goldman Sachs, said on the company’s podcast in June that most of the economic value created by AI investment so far had flowed to chipmakers. During previous technological innovation cycles, semiconductor manufacturers and their customers generally expanded together, but in the current AI cycle, chip companies have generated exceptional profits while businesses higher in the supply chain have yet to demonstrate comparable financial returns.

“Companies ultimately invest in order to make money, and at some point these investments must generate real profits,” Covello said. “That has not yet been demonstrated sufficiently, and it remains the greatest challenge confronting the AI industry.” Alphabet’s negative free cash flow therefore carries significance beyond the company itself. It illustrates a broader tension across the technology industry: revenue and operating earnings may continue to grow, but the amount of capital required to remain competitive in AI is rising even faster, leaving investors uncertain over when the infrastructure race will begin producing sustainable returns.