A.I. Spending No Longer Impresses Wall Street

Earnings season draws a sharper line in the A.I. trade

Wall Street’s latest technology earnings verdict was blunt: It is no longer enough for companies to say they are building for the artificial intelligence future. Investors increasingly want proof that the future is already paying.

That divide came into focus this week as some of the biggest names in tech delivered sharply different messages about the economics of the A.I. boom. Tesla and Alphabet lost hundreds of billions of dollars in market value after signaling that their spending on artificial intelligence would keep climbing. Intel, by contrast, surged after reporting a burst of growth tied directly to the industry’s A.I. buildout. And inside Alphabet’s own results, Google Cloud emerged as a counterweight to investor anxiety, showing that at least some customers are spending aggressively right now.

The market reaction underscored a new phase in the A.I. era. For much of the past two years, simply being associated with artificial intelligence was enough to command investor enthusiasm. Now, as corporate budgets swell and the costs of data centers, chips and computing power mount, investors are sorting companies into two camps: those monetizing demand today, and those asking shareholders to wait.

Tesla and Alphabet face the cost of ambition

The selling in Tesla and Alphabet followed a similar logic, even if the businesses are very different.

Tesla told investors that capital expenditures would continue rising over the next two to three years as it pours money into robotaxis, A.I. computing infrastructure and Optimus, its humanoid robot effort. The company also reported negative free cash flow, reinforcing concerns that its most ambitious projects remain expensive bets rather than near-term profit engines.

Alphabet delivered strong growth in key areas, but investors fixated on a higher spending plan. The company raised its 2026 capital expenditure forecast to a range of $195 billion to $205 billion, a striking figure even by megacap standards. The increase renewed a broader concern hanging over the sector this year: whether the largest tech companies will have to keep escalating A.I. budgets faster than the returns become visible.

That tension has become one of the defining questions of 2026. The industry’s leaders are racing to build more computing capacity, expand data centers and secure enough chips to serve both internal A.I. products and enterprise customers. But every additional dollar of capital spending raises the bar for future revenue, while putting present pressure on margins, free cash flow and valuation.

For Alphabet, the question is not whether demand exists. It is whether demand can scale fast enough to justify a spending plan above $195 billion without weighing on cash generation for longer than investors are willing to tolerate.

For Tesla, the skepticism is even more acute. Investors have heard the long-term thesis around autonomy and robotics for years. What they still lack is firm evidence that robotaxis, Optimus and related A.I. infrastructure will produce returns commensurate with several more years of elevated spending.

Intel becomes an unexpected A.I. winner

If Tesla and Alphabet embodied the market’s impatience, Intel represented the other side of the trade.

The chipmaker reported second-quarter revenue of $16.1 billion, up 25 percent from a year earlier, its fastest revenue growth in nearly 15 years. Its data center and A.I. unit rose 59 percent to $6.3 billion, and its forecast for the current quarter topped Wall Street expectations. Shares climbed sharply as investors embraced the idea that Intel was becoming a more immediate beneficiary of the A.I. spending wave.

The results also strengthened the case that Intel’s turnaround under Chief Executive Lip-Bu Tan is gaining credibility. In a little over a year, Mr. Tan has sought to stabilize a company that had spent years losing ground to rivals and struggling to define its place in the industry’s next chapter. This quarter suggested that the A.I. boom is opening more lanes for Intel than skeptics had assumed.

Company executives said demand is strengthening not just for traditional server chips, but also for custom A.I.-related silicon, advanced packaging and foundry capacity. That is significant because it suggests the A.I. buildout is broadening beyond the most visible winners and spreading across more of the hardware supply chain.

Still, a question remains: whether Intel’s momentum marks a durable reset or a cyclical burst fueled by today’s infrastructure rush. Investors cheered the numbers, but sustaining that confidence will require showing that the company can convert this period of intense demand into a lasting competitive position.

Google Cloud offers evidence of real demand

Within Alphabet’s earnings, one business told a much more encouraging story.

Google Cloud posted revenue of $24.8 billion, up 82 percent from a year earlier, far exceeding expectations. Thomas Kurian, the division’s chief executive, said existing customers are spending roughly 50 percent more than the levels they had already committed to.

That matters because it offers one of the clearest signs yet that enterprises are not merely experimenting with A.I.; many are increasing spending in ways that directly lift revenue for cloud providers. In a market increasingly worried that infrastructure investment may be outrunning customer adoption, Google Cloud’s quarter suggested that at least some of the demand is tangible, urgent and expanding.

The cloud business has become central to the investment case for Alphabet. Search remains immensely profitable, but cloud is where the company is proving that its A.I. tools and infrastructure can translate into fast-growing commercial demand. The stronger that business becomes, the easier it is for Alphabet to argue that today’s spending surge is laying the groundwork for durable revenue growth rather than simply inflating costs.

Why the distinction matters now

This earnings season is shaping up as a referendum not on whether A.I. will matter, but on when it will pay.

That distinction is important because the spending commitments are no longer abstract. Companies are committing tens of billions of dollars to data centers, chips, networking equipment and software infrastructure, often with only partial visibility into the pace of returns. Investors, in turn, are becoming less willing to award blanket premiums for A.I. exposure.

The reaction to this week’s results suggests a more disciplined market. Intel was rewarded because it showed that the buildout is producing immediate sales growth. Google Cloud’s numbers strengthened the argument that enterprises are already opening their wallets. Tesla and Alphabet, by contrast, were punished not because they lack A.I. potential, but because they are asking investors to absorb a longer and more expensive payoff period.

That is likely to be the central test for the rest of Big Tech as earnings continue. The companies that can demonstrate real demand, rising utilization and improving economics may continue to command investor support. Those that mainly offer bigger budgets and longer timelines may face a harsher reception.

For now, Wall Street’s message is clear: In the A.I. race, spending alone no longer wins points. Revenue does.