The US economy’s risky dependance on the AI boom

This article summarizes an episode of WSJ Podcast’s video series featuring Ajay Rajadhyaksha, global chairman of research at Barclays.

Photo credit: Shutterstock

According to Ajay Rajadhyaksha, global chairman of research at Barclays, the US economy is leaning heavily on the current surge in AI spending. If optimism fades, a recession could happen, showing a weakness in the US job market.

The industrial spending boom has one cause

The United States is in its largest period of industrial investment since World War II. But unlike past periods that were driven by many industries, this one is mostly powered by one type of technology.

The spending is so big that it overshadows other industries
Rajadhyaksha estimates, “Just AI spending added 1% to US growth this year. The reason why the US economy shrugged off the trade war so effortlessly: you don’t really care whether General Motors is building a new factory or not if Google comes out and says [it’s] going to spend [US$90 billion] more this year.”

This weakness also affects the stock market
This focus is not just happening in industrial spending. The same thing is happening in the stock market, which makes people feel richer and spend more, all depending on a few companies.

He explains, “Virtually all of the support for the equity market has come from AI-heavy names. They have driven about 80% to 90% of the equity market’s returns, and that has supported the wealth effect, which has supported consumption.”

An uncertain story is holding up the economy

Having so much depend on one area creates a big risk. The whole system now depends on investors staying confident in the AI story.

If that confidence fades, the results will be bad
“If the AI narrative hits a wall,” Rajadhyaksha warns, “there’s no question in my mind that the US economy starts to go perilously close to recession in 2026. AI is a lynchpin.”

Problems could appear first with private companies
While large public tech companies can pay for their AI projects with their own profits, there is more pressure on the startups with a lot of investor money.

He notes, “OpenAI, Anthropic, Perplexity. They have to show me the money. They have to show me the revenues… that’s where the market is likely to be more skeptical, [and] likely to be willing to not give them as much time.”

The threat to jobs is different than it seems

Despite headlines about big layoffs, history suggests job losses are rarely the immediate threat while the economy is still growing.

Rajadhyaksha says, “You almost never, no matter how game-changing a technology there is, you almost never see net job losses coming from it in an expansion.”

The danger appears after the economy slows down
AI impacts on the workforce are subtle. Instead of mass layoffs, companies slowly hire fewer people, which keeps wages from growing.

Who will make the money in the end?

Smart investors may look to other countries