AI bubble concerns grow as handful of companies do all the stock market work

AI investments are driving U.S. economic stability despite President Trump’s tariff and immigration policies creating business uncertainty. Deutsche Bank analysts warn that without tech-related spending, the U.S. “would be close to, or in, recession this year,” but growing concerns about AI’s actual utility could threaten this economic lifeline.

The big picture: Seven major tech companies heavily invested in AI are single-handedly pulling the broader S&P 500 forward, with massive infrastructure investments like the $500 billion Stargate program and Nvidia’s $100 billion OpenAI commitment keeping the economy afloat.

• Nvidia recently became the first U.S. company to hit $4 trillion in market value, followed by Microsoft, both driven largely by AI enthusiasm.
• Google’s Alphabet and Meta have also significantly increased their AI investments and commitments.

Why this matters: Early signs suggest AI adoption may be slowing as companies discover the technology’s limitations, potentially creating a bubble similar to the late 1990s dot-com crash.

• “The reason people are worried about an AI bubble is because seven companies are pulling more than 400 others forward,” Campbell Harvey, a Duke University finance professor, told Al Jazeera.

Warning signs emerging: Major corporations are beginning to reverse their AI implementations after finding the technology less effective than hoped.

• IBM and Klarna initially cut thousands of customer service jobs to replace them with AI, only to start reversing course when they found the technology couldn’t match human workers.
• An MIT report found that 95 percent of companies adopting AI are not achieving significant revenue acceleration from it.
• U.S. Census Bureau data shows AI adoption by large companies has started slowing recently.

What the experts are saying: Academics warn that integrating AI into existing workflows is proving more challenging than anticipated.

• “It turns out, however, that integrating generative AI, in particular, into existing workflows in significantly useful ways is harder than people thought,” said Cal Newport, a Georgetown University computer science professor.
• Newport notes the underlying AI models are “too unreliable” to successfully automate jobs, and rapid job displacement “has simply not come true.”

By the numbers: A Stanford study found entry-level jobs in customer service, accounting, and software development have decreased by 13 percent since 2022 due to AI adoption at large companies.

The bottom line: Unlike the dot-com bubble that left behind useful infrastructure, AI hasn’t yet delivered broad-based productivity improvements that struggling economies need.

• Carl Frey, an Oxford University associate professor, warns: “AI hasn’t yet delivered a clear, broad-based productivity boost—precisely what our stagnating economies need.”
• If the AI bubble bursts without creating lasting economic value, it could cause significant damage to the U.S. economy that currently depends on AI investment for stability.