AI Bubble Has "Less Than a Year Left," Warns Economist Who Predicted the 2008 Crash

Economist Steve Keen, who foresaw the 2008 financial crisis early on, warns that the AI bubble surrounding OpenAI, Nvidia, and Microsoft could burst within a year due to sky-high circular investments.

The AI bubble is back in the spotlight after economist Steve Keen — known as one of the few who saw the 2008 financial crisis coming in time — warned that today's artificial intelligence bubble has "less than a year left." His comments, made to Kitco News, cut to the heart of a debate that major players like OpenAI, Nvidia, and Microsoft can no longer easily brush aside: is too much being invested in AI infrastructure relative to actual revenue?

Who is Steve Keen, and why does his warning matter?

In 2010, readers of the Real-World Economics Review voted Steve Keen the economist who most clearly and most early warned of the 2008 credit crisis. That track record lends extra weight to his latest prediction. According to Keen, today's AI hype shows strong parallels with historical bubbles, such as the nineteenth-century railway bubble: massive capital injections into infrastructure, fueled by expectations of future profits that far outpace reality. Keen estimates that AI companies' actual, recurring revenue currently amounts to only about a fifth of what is being spent on data centers, chips, and computing power.

Circular deals between OpenAI, Nvidia, and Microsoft

A key part of the concern involves so-called circular financing among the biggest AI players. Nvidia invests in OpenAI, which then spends enormous sums on Nvidia chips and Microsoft cloud capacity, while Microsoft in turn depends on that same OpenAI for its AI products. Analysts compare this pattern to the financing structures that preceded the dot-com crash: when suppliers become increasingly dependent for revenue on customers who are themselves dependent on ongoing capital inflows, investor confidence can turn quickly. According to market estimates, OpenAI and Anthropic together have roughly $1.1 trillion in compute-related commitments through 2030, while their combined annual revenue in 2025 was only around $17 billion. That gap between promised spending and actual income is fueling fears that the sector is building on a foundation of optimism rather than hard numbers.

Strong results, but growing doubt

Notably, concerns about an AI bubble are coinciding with sky-high quarterly earnings. Nvidia recently reported a net profit of $26.4 billion for the second quarter of 2026, a 59 percent increase year-over-year, with revenue growing 56 percent to $46.7 billion. Yet commentators like Jim Cramer warn that reports of Nvidia partly financing OpenAI's data center expansion evoke memories of the financing structures seen just before the dot-com crash. Hyperscalers such as Microsoft, Google, and Meta collectively raised more than $244 billion in capital in 2026 alone, as they can no longer fully fund their data center investments from their own cash flows.

What does this mean for the future of AI?

Whether Keen's prediction of "less than a year" will come true remains uncertain. Similar warnings of an impending crash have circulated for some time, and so far chipmakers like Nvidia continue to post impressive revenue growth. Still, a growing consensus holds that the current investment wave, with its towering valuations and interdependencies among a handful of tech giants, is vulnerable to a correction once investors lose patience with the gap between promise and profit. For anyone looking to put the development of artificial intelligence into perspective, it's worth revisiting the history of artificial intelligence: earlier "AI winters" arose precisely when expectations outran technological reality. At the same time, a look at current AI applications shows that the underlying technology is indeed creating value, even if the financial structure around it remains fragile.

Conclusion

Steve Keen's warning has sharply reignited the debate over a possible AI bubble, right as OpenAI, Nvidia, and Microsoft keep reporting record after record. Whether the bubble bursts within a year or not, the gap between investment and actual revenue remains a risk that investors, companies, and policymakers can no longer afford to ignore. Curious how these developments will unfold? Follow more AI news and dive deeper into the background of this fast-changing sector through our knowledge base.

Kitco News

---

Source: Kitco News