The AI Boom Has a Global Money Problem

This story was featured in the Financial Poise Weekly newsletter.

Read the full newsletter for more of the week’s top stories.

As the world pours trillions of dollars into AI, economists are getting an uncomfortable case of déjà vu. Why? Because the way all that money is moving around the globe is beginning to resemble the years leading up to the 2008 financial crisis. And we all remember how that ended.

According to Shawn Donnan, Estelle Ou, and Jorgelina Do Rosario at Bloomberg:

• The world is once again seeing enormous financial imbalances. Countries like China and Japan are accumulating savings while the US borrows heavily. According to IMF projections, global excess savings could reach $1.95 trillion by the end of 2026. That’s a third more than the comparable period before the 2008 crisis.

But things are different this time compared to 2008.

• Last time around, much of that money flowed into US government bonds. Today, more of it is pouring into US stocks, corporate bonds, and AI infrastructure. Bloomberg Economics estimates foreign investors held roughly $4.9 trillion in US-listed AI-related stocks at the end of June.
• And the spending is staggering. One economist estimates that AI infrastructure investment could reach nearly $9 trillion by 2032, surpassing even the 19th-century railroad boom in terms of GDP.

So, what’s the risk?

• If AI fails to deliver the profits investors expect, the fallout could spread far beyond Silicon Valley. Foreign investors, pension funds, and economies increasingly dependent on AI could all feel the pain. And with more money controlled by private investors rather than central banks, a sudden sell-off could make the shock even worse.

Our take? We’ve raised concerns about the circular nature of AI financing before (check out our previous discussion here). Now, we’re seeing just how much bigger the money loop has become. Foreign capital is pouring into the US AI boom, while the countries and investors supplying that money are becoming increasingly dependent on its success.

Does that mean we’re headed for another 2008? Not necessarily. AI is, after all, a real technology with real commercial potential. But even a technological revolution can become an investment bubble when expectations outpace reality. Besides that, if a comparison is to be made, the better one is to the bursting of the internet starting in 2000. Why, you may ask?

• Real Technology vs. Speculative Assets: Unlike the complex financial derivatives and housing market failures that triggered the 2008 crisis, both the internet in 2000 and AI today represent genuine technological revolutions with real, transformative commercial potential.
• Investment Bubbles Driven by Expectations: Even revolutionary technologies can experience severe market corrections when investor enthusiasm and capital inflows outpace immediate commercial reality.
• Massive Infrastructure Spending: The staggering scale of capital pouring into AI infrastructure– projected to reach $9 trillion by 2032– mirrors the massive overinvestment in fiber optics, telecommunications, and dot-com startups during the late-1990s boom before valuations corrected.

Regardless, much of the world’s money is riding on AI’s promise to transform productivity, which leads us to our next story… Are Businesses Seeing Productivity Gains from AI?