The IMF’s chief economist, Pierre-Olivier Gourinchas, said the ongoing U.S. AI investment surge could face a market correction similar to the dot-com bust, but is unlikely to destabilize the global economy due to its low reliance on debt financing.
“Unlike the 2008 crisis, this is not debt-driven,” Gourinchas said, noting that tech firms are funding AI growth with cash, not leverage. “If there is a market correction, some shareholders may lose out, but it won’t spread through the financial system.”
The IMF compared the current wave of AI enthusiasm to the internet bubble of the 1990s, when expectations for transformative technology led to inflated valuations that later crashed. However, today’s AI investment remains modest — about 0.4% of U.S. GDP since 2022, versus 1.2% during the dot-com era.
Gourinchas said a potential AI downturn could still impact market sentiment and asset prices, especially in non-bank financial sectors, but not at a systemic level. The IMF’s World Economic Outlook highlighted AI spending as a short-term driver of growth, while warning that it may be fueling inflation before productivity benefits appear.
The IMF forecasts U.S. inflation will decline more slowly than expected, to 2.7% in 2025 and 2.4% in 2026, partly due to tariff-related price pressures and a smaller labor force from reduced immigration. Gourinchas noted that despite political claims, U.S. firms — not exporters — are largely absorbing the cost of Trump’s tariffs.
