Artificial intelligence is reshaping the world economy at a speed that rivals the steam engine in the nineteenth century. Yet the scale of investment and the debt being piled up to fund the AI boom raise fresh concerns about a possible financial shock.
The AI investment surge
US tech giants that run massive data centres, known as hyperscalers, are expected to spend between $600 billion and $800 billion on capital projects this year. Roughly 70 % of the AI revenue generated by these hyperscalers comes from just two firms, OpenAI and Anthropic. While Anthropic is currently expanding, OpenAI has missed its own revenue targets and is experiencing a notable departure of senior staff, a sign that the company may be facing deeper problems.
Debt-fuelled risk
Wall Street veteran Steve Eisman, who famously predicted the 2008 crash, has cautioned that a collapse of a major AI player could set off a chain reaction of defaults. The United States already carries a public debt load of about $40 trillion, a level that makes any large-scale bailout increasingly fraught.
Investors are already showing unease. The world's largest sovereign wealth fund, Norway's, has announced plans to reduce its exposure to US government bonds, signalling a broader wariness about the intertwining of public and private debt.
Potential spill-over to Europe
Europe could feel the impact in two ways. First, a US-centric AI bust could depress global stock markets and weaken the dollar, making American assets less valuable for European investors. Second, a stronger euro that often follows a weaker dollar would hurt the export-driven economies of the eurozone, from Germany to Spain.
What could happen next
If a leading AI firm were to fail, investors and lenders might demand immediate repayment, putting pressure on the US Treasury. The Federal Reserve could respond by buying government debt and cutting rates, actions that risk fuelling inflation and further eroding confidence in the dollar. Such a scenario could quickly evolve into a worldwide financial shock, echoing the chain reactions that followed the sub-prime crisis of 2008.
While the transformative potential of AI, from legal services to manufacturing, remains clear, the financing story is more complex. The technology itself is unlikely to be the cause of a crash; rather, it could be the excessive optimism and debt built around it that triggers a crisis.
Policymakers, investors and businesses would do well to prepare for this possibility, even as they continue to harness AI's productivity gains.