Big Tech’s AI Spending Surge Drains Cash Even as Revenue Climbs

Alphabet swings to a $5.9 billion free-cash-flow deficit in the second quarter

Shares slide as AI investment costs eclipse 82% cloud growth

Big Tech’s race to lead in artificial intelligence is turning into a drain on cash flow. AI-related revenue is rising quickly, but spending on data centers, semiconductors and power infrastructure is increasing even faster.

Alphabet shares fell $24.40, or 7.13%, to $317.69 on the New York Stock Exchange on July 23 after the company reported second-quarter results. The earnings themselves were not weak. Revenue continued to grow at a double-digit pace, and Google Cloud revenue jumped 82% from a year earlier. Demand for AI infrastructure and AI solutions drove that growth.

The problem was cash flow. Alphabet posted a $5.9 billion free-cash-flow deficit in the second quarter. The surge in AI infrastructure spending meant cash outflows for investment exceeded cash generated. Alphabet also raised its capital-expenditure outlook for this year to $19.5 billion to $20.5 billion.

That is where Big Tech’s dilemma begins. AI is generating revenue. Cloud usage is rising, and corporate adoption of AI is accelerating. But running AI services requires massive data centers, high-performance semiconductors and power equipment. Even as revenue grows, investment spending is going out faster.

Tesla showed a similar pattern. Its second-quarter capital expenditures surged 142% from a year earlier to $5.8 billion. Free cash flow swung to a $1.1 billion deficit. Tesla is expanding beyond electric vehicles into robotaxis, humanoid robots and AI infrastructure, but those investment costs are weighing on near-term performance.

Oracle is another example of the burden of AI investment. In fiscal 2026, cloud revenue increased sharply, but free cash flow showed a $23.7 billion deficit. Expanded investment in AI data centers drove the cash outflow. Order backlog and cloud growth prospects have improved, but investors have started asking not only how much money companies can make, but also how much they must spend to earn it.

In the past, Big Tech could fund new businesses with the huge cash generated by search, software, advertising and e-commerce. The AI race is different. The scale of investment is far larger, and the payback period is longer. Some companies are even turning to debt issuance and stock sales to fund the spending.

Investor views are divided. Optimists see AI infrastructure spending as a necessary cost for long-term growth. The logic is that companies that fail to pour money into data centers and semiconductors now could fall behind in the AI market later. Demand for AI is already showing up in cloud revenue and growth in enterprise services.

More cautious investors worry that spending is rising faster than monetization. As capital expenditures increase, depreciation and operating costs also climb. Even if AI revenue rises, the improvement in cash flow shareholders are waiting for could be delayed if that revenue does not outpace investment costs, operating expenses and depreciation.

Lee Song-ryeol, Hankyung.com reporter yisr0203@hankyung.com