Big Tech’s latest earnings season has turned into a referendum on artificial intelligence spending, and investors are voting with sell orders. Amazon, Google and Microsoft are set to lose a combined $900 billion in market value after filing quarterly results over the past week, according to the Financial Times, as Wall Street recoils from a wave of planned spending that has pushed the AI boom into bubble-fear territory. The immediate trigger is not weak sales. It is the sheer size of the cash burn now attached to the race for AI infrastructure.
The latest numbers are staggering even by Big Tech standards. The FT says Amazon, Google, Microsoft and Meta are planning a combined $660 billion in capital expenditure in 2026, up 60% from $410 billion in 2025 and 165% above $245 billion in 2024. That spending trajectory has become the market’s new stress test. After months of rewarding companies for proving they can monetize AI, investors are now asking a more uncomfortable question: how long can profits and cash flow absorb this level of investment before returns show up?
The answer, at least for now, is not reassuring. Amazon said 2026 capital expenditure will reach $200 billion, about $50 billion more than expected, and the stock fell 8% in premarket trading Friday, according to the FT. Microsoft has dropped 18% since reporting on Wednesday, even though cloud revenue rose 26% to $51.5 billion. The catch is that data-center spending surged 66%, reinforcing the sense that the company is funding an arms race whose payoff is still hard to pin down.
Investor anxiety is not limited to the size of the bill. Microsoft also disclosed that 45% of its $625 billion future cloud-contract book comes from OpenAI, sharpening concerns about concentration risk inside one of the biggest AI partnerships in the industry. At the same time, the company is being measured against expectations that have moved far faster than its ability to explain the economics. Revenue is growing. So is spending. That combination is no longer enough to keep the market calm.
The same tension is now hanging over Alphabet. Google’s parent surpassed $400 billion in annual revenue and $132 billion in profit in 2025, yet it still plans to double capital expenditure to $185 billion. That is a striking contrast: a company generating enormous profits is still being asked to spend at a pace that makes investors wonder whether the race for AI infrastructure is becoming self-defeating. Strong earnings did not end the selloff. If anything, they reminded the market that even the most profitable platform companies are now tied to an increasingly expensive investment cycle.
Meta is caught in the same trap. The company said capex would double to $135 billion, and its shares initially rose 10% before giving back the gains. That reversal captures the mood shift better than any model does. Investors are not rejecting AI outright. They are rejecting open-ended spending without a clear timeline for returns. In a market that spent much of the year treating AI as the safest growth theme in tech, the mood has become much more selective, and much more suspicious.
The loss of confidence has spread beyond the software giants. Oracle fell 18% over five days after the OpenAI and Nvidia-backed $100 billion investment and infrastructure deal did not go ahead, according to the FT. That pulled another layer of momentum out of the AI trade, suggesting investors are now questioning not only the hyperscalers’ budgets but also the broader financing structure around the buildout. When deals fall through and capex rises anyway, the story starts to look less like a growth cycle and more like a funding gap.
There is a simple reason the selloff has been so fast. The market can tolerate high spending when the payoff is visible and near-term. It gets nervous when spending rises faster than evidence of return. Jim Tierney, head of concentrated US growth fund at AllianceBernstein, put it bluntly: “The capex is breathtaking.” That line lands because it describes more than a quarterly trend. It describes a whole phase of the market in which the biggest companies are acting like infrastructure builders, while investors are still trying to value them like software platforms.
Anna Nunoo, senior analyst at AllianceBernstein, said, “The onus is on Microsoft and Amazon to prove out the attractive returns on all the spending.” That is the central issue now. Amazon can argue that cloud demand justifies the outlays. Microsoft can point to cloud growth and customer demand. Alphabet can cite its huge revenue and profit base. Meta can say the long-term opportunity in AI is too large to ignore. But none of those arguments remove the burden of proof. In this market, proof means visible returns, not just bigger budgets.
Brent Thill, an analyst at Jefferies, summed up the change in tone: “AI bubble fears are settling back in. Investors are in a mini timeout around tech, and nothing the companies say fundamentally matters.” The wording may sound extreme, but the stock action backs up the sentiment. Amazon, Microsoft and Alphabet are all being judged not on whether they can keep growing, but on whether the cost of staying in the race is getting out of hand. The market is no longer asking who can spend the most. It is asking who can spend the most without destroying returns.
One company offered a sharp contrast. Apple was the only winner in the group, up 7.5% after reporting record quarterly revenue of $144 billion. That does not mean Apple is suddenly the market’s favorite AI play. It means investors currently prefer a business model that is easier to forecast, with less immediate capital intensity and less exposure to the infrastructure arms race. In a week when the rest of Big Tech was being punished for planning too much spending, Apple’s cleaner earnings profile looked almost conservative.
The broader market is feeling the stress, too. The Nasdaq Composite is down 4% over the past five days and is on course for its worst week since April. That decline matters because it shows the AI selloff is not isolated to a few headline names. It is feeding through the entire tech complex, dragging index performance and investor sentiment with it. When the biggest companies in the market start to wobble at once, the rest of the sector usually loses some of its cushion.
Nvidia is now the next catalyst, and the market will be watching closely for any sign that the chip leader can keep the AI trade from unraveling further. But for now, the message from Big Tech earnings is clear: the AI boom is not over, yet the price tag has become impossible to ignore. Investors wanted proof that spending would eventually pay off. Instead, they got a bill.
