The AI Spending Boom Just Hit New Records: Inside Big Tech's ~$700 Billion Bet

The scale of the AI build-out got hard to ignore this week. In the space of a few days, Google posted record capital spending that pushed its cash flow negative, and chipmaker AMD agreed to pour up to $5 billion into AI company Anthropic. Two very different moves, one very clear message: the money going into AI infrastructure is now measured in the hundreds of billions — and it is still climbing.

Put the pieces together and a trend snaps into focus. Across the largest tech companies, 2026 spending on AI infrastructure — the data centers, chips, and power behind every chatbot and cloud service — is running toward an estimated $700 billion for the year, sharply higher than 2025, according to widely cited industry estimates. This is the story behind a lot of the market's biggest moves right now, so it is worth understanding what is actually being spent, why, and what to watch.

This article is for information and education only and is not financial advice. Nothing here is a recommendation to buy, sell, or hold any security. Company figures are drawn from recent reports and estimates and may be revised.

The boom, by the numbers

Three recent data points capture the pace:

• ~$700 billion: the estimated total 2026 AI-infrastructure spending across the big "hyperscalers" — companies like Google, Microsoft, Amazon, and Meta — a sharp jump from the prior year, per industry estimates.
• ~$44.9 billion: Google's capital spending in a single quarter (Q2 2026), roughly double a year earlier. It was enough to flip the company's free cash flow negative for the first time in about a decade, and it lifted full-year spending guidance to roughly $195–205 billion.
• Up to $5 billion: what AMD agreed to invest in AI company Anthropic, alongside a deal to supply it with tens of billions of dollars of AI server chips.
• ~$35 billion: the largest private credit deal on record, arranged by investment firms to help finance Anthropic's computing expansion — a sign of how much of the build-out is now funded by debt.

Different companies, different moves — but all pointing the same way: the companies building the most AI computing capacity now are positioning to serve fast-growing demand.

What everyone is actually buying

The spending goes into a few big buckets: data centers, the specialized chips that run AI models, and the electricity to power both. Demand is the reason. AI services are increasingly showing up as real revenue — Google's cloud business, for instance, grew about 82% year over year last quarter — and companies are racing to add enough capacity to serve it. In an arms race like this, spending less than a rival can look riskier than spending more, which is part of why the numbers keep climbing.

The circular twist: when a chipmaker invests in its customer

The AMD-Anthropic deal is a good window into a pattern that has become one of the defining features of this boom. AMD is not just selling chips to Anthropic; it is also investing in it. That combination — a supplier putting money into the very customer that will buy its products — is common enough in AI right now that analysts have a name for it: circular financing (or vendor financing). A leading chipmaker has made similar moves, taking stakes in the AI labs that are among its biggest buyers.

It reflects a broader pattern in the AI industry: partners increasingly investing in one another to build capacity together. The logic is that the investment helps the customer grow, and a bigger customer buys more chips — deepening the partnership on both sides. It is a useful concept to understand, because these kinds of tie-ups are becoming more common as companies line up the pieces they need to expand.

One practical note: Anthropic is a private company, so it is not something an everyday investor can buy shares in directly. The publicly traded names in this story are the chipmakers and cloud giants — but understanding the deal is about reading the trend, not acting on it. For the broader idea of investing in fast-growing private companies, see Finelo's explainer on venture capital.

The cost side: how "record growth" and "negative cash flow" share a quarter

Here is the part that trips up a lot of newer investors. How can a company grow revenue strongly and still report negative free cash flow in the same quarter? Because free cash flow is roughly the cash a business generates from operations minus what it spends on long-term assets (capital expenditure). When a company chooses to spend more on building — as Google did with its record AI capex — than its operations throw off, free cash flow turns negative even as the business booms.

That negative number is not automatically a warning sign; for a highly profitable company it usually reflects a deliberate bet on future growth rather than distress. But it does mean less cash left over for buffers, buybacks, or dividends, and it puts the spotlight on one question: will the investment earn an adequate return? The market's reaction to Google's quarter — strong results, yet a dip in the shares — was investors weighing exactly that. For how this fits with other financial measures, see Finelo's guide to EBITDA.