Alphabet (GOOGL) ended Thursday at the center of the market’s worst selloff in a month after raising its 2026 capital-spending range to $195 billion to $205 billion. The S&P 500 closed down 1.2%, while the Nasdaq dropped 2.2%, as Alphabet sold off sharply and Tesla fell even harder.
The immediate question is no longer whether companies can find demand for artificial intelligence. Alphabet’s Cloud revenue grew 82%, which is clear evidence that demand exists. The harder question is whether Big Tech can fund this buildout from its own operations before the cost of power, equipment, and capital makes the payback too slow.
Cash, not demand, is now the dispute.
A spending plan that changed the frame
Alphabet reported $44.9 billion of capital expenditures in its latest quarter and its first recorded cash burn. Capital spending is money used to build long-lived assets, such as data centers and computing equipment. Cash burn means the company spent more cash than it generated over the period.
That combination carries more weight than a routine increase in a technology budget. Alphabet has one of the world’s strongest internet businesses, plus a fast-growing cloud operation. If its AI investment can push quarterly cash flow below zero, smaller or less profitable companies have a much thinner margin for error.
Thursday’s decline had more than one driver. Brent crude briefly moved above $100 a barrel during the session, adding fresh concern that energy costs could lift inflation and delay interest-rate cuts. But Alphabet’s spending plan gave investors a specific reason to reconsider the value of profits expected far in the future.
That is the key shift. A large AI budget can signal competitive ambition and real customer demand. It can also signal that the company must spend heavily for years before shareholders see a matching lift in cash returns. Both can be true at once.
The market has become less willing to treat those two ideas as the same thing.
Alphabet has not said that AI demand is weak. Its Cloud growth points in the other direction. The issue is the sequence of events: cash leaves the business when facilities and equipment are built, while the revenue and profit from that capacity may arrive over a much longer period.
For years, investors have broadly rewarded large technology companies for spending ahead of demand. That approach worked when borrowing costs were low, energy was cheaper, and the largest platforms could show growing profit alongside their investments. The test gets tougher when each new dollar of infrastructure has a higher cost and a longer route to a return.
Revenue growth does not settle the return question
Cloud revenue growth of 82% is the strongest part of Alphabet’s case. It shows customers are paying for computing capacity and AI services. It also gives management a credible reason to add more data-center capacity rather than wait for demand to overwhelm existing systems.
Yet revenue growth alone cannot show whether the new investment will earn an adequate return. Data centers require up-front cash. The equipment inside them loses value over time, a cost known as depreciation. The faster hardware changes, the greater the risk that older equipment becomes less useful before it has produced enough profit.
AI equipment may face that problem more sharply than older technology cycles.
Alphabet therefore needs to show more than higher cloud sales. It needs to show that added AI usage produces enough profit to cover the equipment, the power required to run it, and the ongoing cost of expanding capacity. A company can post strong sales growth while still weakening its cash economics if spending rises faster than the profit from those sales.
This is why the annual capital-spending range drew such attention. It changes the size of the bet. Investors now have to judge whether a far larger infrastructure base will become a durable earning asset or a costly race to keep up with rivals.
The favorable case is real. Data-center spending often has a high fixed cost at the start. Once the capacity is in place, more customer usage can improve returns because it spreads those fixed costs across a larger revenue base. If AI demand converts quickly into higher-value cloud workloads, Alphabet could eventually earn more from infrastructure that looks expensive today.
That outcome depends on timing and margins. Faster usage helps. So does a product mix that earns enough per unit of computing to justify the equipment bill. Neither is guaranteed simply because demand is rising.
The countercase is concrete: Alphabet could build capacity ahead of demand, or the demand could prove less profitable than investors assume. In that case, Cloud revenue could remain strong while depreciation, power needs, and cash demands weigh on returns. The observable warning sign would be continued heavy spending without clearer evidence that incremental cloud revenue is turning into stronger cash generation.
Oil raised the hurdle for distant profits
Oil’s move during Thursday’s session matters because AI data centers consume substantial power. Brent above $100 was brief, and one trading day does not establish a lasting inflation trend. Still, it reminded investors that the AI buildout relies on more than chips, software, and customer demand.
Power has a price. So does time.
A sustained rise in energy costs can spread through transport, industrial activity, and electricity markets. It may also make central banks more cautious about lowering interest rates. Higher rates raise the return investors demand from projects whose payoff lies years ahead.
That does not affect every business equally. Companies with current cash flow tied to energy production may have more financial flexibility if commodity prices remain high. Their operations can generate more cash today, which can support investment, debt reduction, or shareholder returns.
High-valuation growth companies face a different math problem. Their value often rests more heavily on earnings expected well into the future. When interest rates stay elevated, those future earnings are worth less in present terms. A long AI payback period then becomes harder to support, especially if energy costs also rise.
Alphabet is not an energy company, and higher oil prices do not automatically translate into its power costs. The read-through is broader: a higher-cost economic setting leaves less room for infrastructure projects that need years to prove their value. Investors will likely focus more closely on cash conversion, which measures how well reported earnings become actual cash after spending needs.
The important signal is not whether oil stays above a single threshold. It is whether energy pressure begins to feed into inflation concerns and keeps the cost of capital higher for longer. That would raise the burden of proof for every company asking investors to wait for AI returns.
Tesla showed the same discipline
Tesla (TSLA) fell roughly 14.5% Thursday after reporting weaker profit and outlining a heavy investment plan. Tesla and Alphabet have very different businesses, but both now face the same basic challenge: show that large spending commitments will turn into earnings fast enough.
Tesla’s autonomy work has a less certain path to revenue than a standard factory expansion. The company must spend before investors can see whether its technology becomes a product or service that customers will buy. Weaker profit leaves less current earnings power to support that wait.
The share-price decline does not prove why every seller acted. It does show the question now facing companies with large future projects. A technical ambition may be necessary to defend a business over time, but it does not remove the need to explain the cash cost and likely return.
That is the common thread between Tesla and Alphabet. The issue is not whether AI infrastructure or autonomy could produce attractive economics. Both could, particularly if demand arrives quickly after the initial investment. The issue is whether the spending curve is rising faster than the profit curve.
Next week’s Big Tech reports are the next test. If other companies pair larger AI budgets with clear signs that revenue and cash returns are keeping pace, Alphabet may look like a company-specific squeeze. If spending rises across the group while the evidence on returns stays thin, Thursday’s selloff may mark a tougher new rule for the AI trade.
