Amazon, Google, Microsoft, and Meta are spending nearly $700 billion on artificial intelligence this year. 

That’s more money than the entire economy of Sweden. And it’s causing a financial crisis that’s making Wall Street question whether the AI boom is sustainable.

Amazon’s free cash flow is going negative. Meta’s is dropping 90%. And the stock market is punishing them for it.

The $700 Billion Question

Amazon announced Thursday it would spend $200 billion in capital expenditures this year.

Google isn’t far behind at $175-185 billion. Microsoft is on pace for $145 billion. Meta plans to spend $115-135 billion.

To put that in perspective: $700 billion is larger than Sweden’s entire GDP. It exceeds the combined capital spending of 21 other major U.S. companies—including Exxon Mobil, Intel, and Walmart—by more than 3.5 times.

This is four companies, all in the same sector, all betting everything on the same technology, all at the same time.

Amazon’s Free Cash Flow Just Went Negative

Free cash flow is the money a company has left over after paying for operations and capital expenditures. It’s the cash available for dividends, stock buybacks, or future investments.

Amazon’s free cash flow is expected to turn negative in 2026. Morgan Stanley analysts project -$17 billion. Bank of America sees -$28 billion.

That means Amazon will be burning cash—spending more than it earns—just to keep up with the AI buildout.

Meta’s free cash flow is expected to drop nearly 90%. Bank of America warns that the four hyperscalers are “reaching a limit to how much AI capex they are willing to fund purely from cash flows.” In 2024, AI spending consumed 76% of their operating cash flow. In 2026, that number hits 94%.

When you’re spending 94% of your cash on a single bet, there’s no room for error.

Amazon Filed It May Need To Raise Money

On Friday, Amazon filed a notice with the SEC saying it may seek to raise equity and debt as its AI buildout continues.

When a company as profitable as Amazon says it might need to raise money, it’s an admission that internal cash generation can’t keep up with spending.

The Stock Market’s Verdict: Punishment

Wall Street’s reaction has been brutal.

Amazon dropped 8% Friday. The stock is down 9% year-to-date. Microsoft fell 11% after earnings and is down 17% for the year. Google dropped 3% after announcing it would nearly double capital spending.

The pattern is clear: the more a company spends on AI, the worse its stock performs.

The one exception? Meta rallied because it’s the only company clearly showing how AI is boosting revenue through ad targeting. Everyone else is asking investors to trust that the spending will pay off eventually.

Meanwhile, the companies selling AI infrastructure are thriving. Nvidia jumped 5% Friday. Broadcom rose 5%. AMD spiked 6%. The companies building AI are getting punished. The companies selling them the picks and shovels are getting rewarded.

Why Are They Spending So Much?

Microsoft CFO Amy Hood explained it bluntly: “We’ve been short [on computing power] now for many quarters. I thought we were going to catch up. We are not. Demand is increasing.”

Microsoft’s backlog has doubled to $625 million, largely thanks to OpenAI. Amazon CEO Andy Jassy cited “strong demand for our existing offerings and seminal opportunities like AI.” Meta CEO Mark Zuckerberg described an “unsatiated appetite for more computing resources.”

These aren’t speculative bets. These are companies responding to real customer demand they can’t currently meet.

But if demand is so strong, why isn’t revenue growing fast enough to fund the buildout?

The Revenue Problem

OpenAI—the clearest success story in AI—ended 2025 with about $20 billion in annual revenue. Microsoft is targeting $25 billion in AI-related revenue by fiscal 2026. Google Cloud revenue grew 48% to $17.7 billion in Q4 2025.

These are real numbers. But the hyperscalers are spending $700 billion to build infrastructure for an industry generating maybe $50-60 billion in total annual revenue.

That’s a 12-to-1 ratio of spending to revenue.

At some point, the math has to work.

Are We In A Bubble?

During Microsoft’s earnings call, an analyst asked: “Are we in a bubble?”

It’s a fair question. The last time we saw this kind of concentrated spending was the dot-com era, when telecom companies spent hundreds of billions building fiber-optic networks. The growth happened, but not fast enough. Most companies went bankrupt. The infrastructure was eventually used—but by companies that bought it for pennies on the dollar.

This time is different. The hyperscalers are profitable businesses with real customers paying for AI services right now. But the scale of spending relative to current revenue is unprecedented.

McKinsey estimates $7 trillion in data center investment will be required by 2030. That’s over four years, from an industry currently generating a fraction of that in revenue.

What Happens Next?

Big Tech is betting that AI demand will continue growing exponentially and that revenue will eventually catch up to spending.

They might be right. AI could genuinely transform every industry, creating trillions in value.

Or they could be overbuilding. Revenue growth could slow. Enterprise customers could decide AI tools don’t deliver enough value. And we could see a painful correction.

For now, investors are voting with their wallets. Their message is clear: prove it pays off, or we’ll keep punishing your stock.

The $700 billion question isn’t whether AI is real. It’s whether the current spending pace is sustainable—and whether the returns will ever justify the cost.

Guardian Financial Publishing, 3571 Far West Blvd · Texas · Austin · 78731 · United StatesNo longer want to receive these emails. Unsubscribe