The IMF’s top economist warns that the AI surge echoes previous bubbles, such as the dotcom bubble, where valuations and paper wealth boost demand before productivity catches up.
The international funding body argues that a correction is plausible, but unlike 2000 or 2008, it’s less likely to be systemic because Big Tech is funding capex with cash, not debt.
Since 2022, AI-related investment has risen by under 0.4% of US GDP, a fraction of the dot-com era’s 1.2% lift between 1995 and 2000. Still, the wealth effect is doing macro work.
The IMF’s latest outlook credits AI spend, a softer-than-feared tariff path, easier financial conditions, and a weaker dollar for propping up US and global growth.
The catch is inflation, with the IMF now contending that as the US CPI eases to only 2.7% in 2025 and 2.4% in 2026, still above the Fed’s 2% target, AI-driven investment and consumption could keep demand aloft while non-tech capex drifts.
Tariff dynamics complicate the mix. Import prices haven’t fallen, and US firms appear to be absorbing costs in margins rather than passing them on, which is a slow, sticky route back to price stability. The IMF also flags lower immigration as a constraint on laborsupply.
What Does This Mean for Me?
If an AI correction arrives, the immediate damage would likely center on equity holders, with second-order risks from a broader risk-off repricing hitting non-bank financials. For investors, the message is simple: treat AI as a growth tailwind with inflation side effects, not a leverage-fueled bubble, yet.