BIS warns: AI bubble could trigger credit crisis like 2008

What happened?

On June 23, 2025, the Bank for International Settlements (BIS), known as the central bank of central banks, published its annual report with an unprecedented warning: a possible burst of the artificial intelligence (AI) investment bubble could affect credit markets with a magnitude comparable to the 2008 financial crisis. The report, cited by The Next Web, lists AI-led risks alongside inflation and fiscal stress as 'pressure points' that 'demand attention.' This warning comes at a time when global AI investment reached $500 billion in 2024, according to OECD data, and is expected to exceed $1 trillion by 2027. The BIS's concern centers on a pattern of 'circular financing': big tech companies invest massively in AI infrastructure (data centers, specialized chips, energy), and in turn, providers of that infrastructure use AI tools to optimize their operations and attract more investment. This cycle can artificially inflate demand and create a speculative bubble similar to the dot-com bubble, but with greater credit leverage.

Why is it important?

The relevance of this warning lies in the BIS's authority as a supervisor of global financial stability. Founded in 1930, the BIS coordinates the world's central banks and has access to data that no other institution possesses. In its report, the BIS notes that 'circular financing' in AI recalls the 'shadow banking' that preceded 2008, where banks lent to each other to inflate mortgage assets. The difference is that now the cycle is faster and more globalized: AI companies issue corporate debt at low rates, and investment funds buy that debt by leveraging themselves in turn. If AI returns fail to meet expectations (for example, if business adoption stalls or restrictive regulations emerge), a sudden capital withdrawal would occur. According to the BIS, total risk exposure could exceed $2 trillion, equivalent to 2% of global GDP. Additionally, the report highlights that the concentration of investment in a few companies (Google, Microsoft, Meta, Amazon, Nvidia) creates systemic risk: if one of them collapses, it would drag down its suppliers and interconnected credit markets.

What consequences will it have?

The consequences of an AI bubble burst would be severe and multifaceted. First, the BIS warns that an abrupt correction could spread to credit markets, as many AI investments are leveraged with debt. For example, data center companies have issued bonds worth over $300 billion in the last three years, and banks have granted syndicated loans for another $200 billion. If the value of underlying assets (servers, patents, startups) collapses, banks would suffer massive losses, triggering a global credit crunch. This would affect businesses and consumers, reducing consumption and investment. Second, the report notes that central banks must prepare to intervene, but traditional tools (lowering interest rates, injecting liquidity) may be insufficient given the speed of digital markets. The 2008 crisis took months to develop; an AI crisis could occur in days due to the automation of algorithmic trading. The BIS recommends creating specific 'capital buffers' for AI exposures and improving oversight of leveraged investment funds. Finally, confidence in technology would be eroded, delaying AI adoption in key sectors like healthcare or transportation, which would have a huge opportunity cost.

What should readers know?

Readers should understand that the BIS warning is not a prediction of an imminent collapse, but a wake-up call about systemic risks. Tech companies that rely on external financing for their AI projects could face difficulties if market sentiment shifts. Investors should diversify and not overexpose themselves to AI-linked assets, especially those with high leverage. Regulators, for their part, need to closely monitor leverage in the sector and establish transparency requirements. The BIS compares the current situation to the dot-com bubble (1995-2000), but notes that credit risk is greater due to today's financial interconnectedness. While in 2000 most dot-com companies had no debt, today big tech firms have solid balance sheets, but AI startups are highly leveraged. Moreover, the speed of innovation and adoption is much higher, amplifying both ups and downs. The key lesson is that technological innovation is not immune to boom-and-bust cycles, and prudence is essential. As the report states: 'History shows that investment bubbles usually end badly. The BIS reminds us that AI is no exception.'

"History shows that investment bubbles usually end badly. The BIS reminds us that AI is no exception," the report states.

In summary, the BIS report is a wake-up call for governments, investors, and companies. AI has transformative potential, but its current financing presents systemic risks that should not be ignored. Prevention and prudent regulation will be key to preventing an AI correction from triggering a global financial crisis.