Andrew Bailey, Governor of the Bank of England, has warned of a sharp downturn in global stock markets if the AI bubble bursts.
Bailey expressed his concern in a letter to G20 finance ministers, in which he addresses the economic consequences if the debt-driven investment boom in AI begins to unravel, writes The Telegraph.
He warned that the soaring levels of debt that have been used to finance AI “could amplify a future market correction”.
A crash among AI companies would cost investors trillions of dollars overnight. This could trigger a dangerous chain reaction if they were then forced to sell other assets. The danger is that the entire stock market would then go into a total meltdown, so that we experience a new and, if possible, far worse financial crisis than in 2007–2008.
The crisis would be an outright catastrophe in a situation where the major Western economies are already characterised by high debt, industrial decline and dismal economic growth. Several countries that were previously economic great powers, such as Germany, are showing clear signs of recession. Bailey expresses himself more mildly in the letter, but between the lines the seriousness of the warning is clearly apparent; it has also been reported by The Guardian and several other British media outlets.
“Recent developments have also made it clear to me that many jurisdictions do not have protocols in place to manage the development, release and deployment of advanced, frontier AI models, increasing risks to the financial sector and other sectors.”
Bailey said that the “frontier” AI models “demonstrate increasingly sophisticated autonomy and problem-solving capabilities, as well as a capacity to pose a threat”.
Such models risk destabilising the “highly interconnected” global financial system through cyber disruptions that “could spread across jurisdictions”.
The risk is amplified by investors borrowing enormous sums to invest in a small number of AI companies and data-centre providers. This pushes the valuations of these companies up to astronomical levels and increases the likelihood that the bubble will eventually burst.
Nvidia, which is worth more than $5.2 trillion, recently raised $500 billion from a consortium of American banks and investors to finance its wave of investment in AI. Its share price has risen by an incredible 850 per cent over the past five years.
Such “overvalued assets” in AI are one factor that could trigger a sharp economic downturn, Bailey warned.
“Markets remain vulnerable to a potentially disorderly correction that could spread across borders. The problem is not merely that investors are borrowing more, but that leverage interacts with high valuations and market concentration.”
Bailey’s warning came as the Labour government unveiled a £100 million fund to support British AI startups. The fact that the authorities are getting involved will be regarded by many as almost a guarantee that the bubble is soon about to burst.
International cooperation is the solution
Alarm bells have also been ringing among prominent technologists in recent weeks, echoing Bailey’s earlier calls for international cooperation to deal with the growing threats from AI. As he told leaders in the City:
“No country can shut itself off from the cross-border nature of the systems that are widespread today.”
It is difficult to see how international cooperation is supposed to have any positive effect. Such cooperation usually ends with politicians setting up committees that are supposed to be able to reduce the threat.
Such committees are often filled with climate hysterics and people who completely lack relevant education, experience and understanding of the international stock markets, not to mention an understanding of the threats and opportunities posed by the drive towards AI.
Last month, a letter was signed by 1,367 researchers and engineers at frontier AI laboratories – mainly OpenAI, Anthropic and Google DeepMind. The letter highlighted the concerns of the engineers who work with the technology every day. They, too, regard global cooperation as a solution.
“There is a real risk that capability development will rapidly accelerate beyond our ability to understand or control the resulting systems.”
These companies, too, regard global cooperation as a solution. In the letter, they called for support from the US government for “an international effort to develop the technical and governance tools needed to deliberately regulate the pace of development of automated artificial intelligence”.
Bailey sent his letter in his capacity as chairman of the Financial Stability Board, which works with central banks and governments around the world. Bailey also warned that AI risked triggering a wave of cyberattacks that could cripple the global economy.
“The risk landscape has been further complicated by the emergence of frontier AI models, which demonstrate increasingly sophisticated capabilities in autonomy and problem-solving, as well as threat capabilities.
Frontier AI may have the ability to materially alter the speed, scale and economic consequences of cybersecurity risk, which could undermine market confidence in the system as a whole.”
A lack of cybersecurity combined with increasingly advanced hacking could pose a major threat to the global financial system, which is very tightly interconnected. In other words: If the bubble bursts in one market, the entire global market will be affected within seconds, and the crisis could be greatly amplified by the panic that then ensues.
Today’s politicians can prolong the crisis through their belief that public support can solve every problem. We saw this after the stock-market crash of the interwar period. Many prominent economists, including Thomas Sowell, believe that the then US President F. D. Roosevelt prolonged the crisis for years with his “New Deal”, which actually increased the consequences of the crisis after it was implemented.
Green New Deal er basert på en løgn om at New Deal var en suksess
Friedrich Hayek was also a strong critic of the New Deal, since he regarded government interference in the economy as a threat to individual freedoms and a dangerous step towards centralised economic planning and totalitarianism.
We saw the same thing during the financial crisis of 2007–2008, when government interference made matters worse. Former President Joe Biden wanted to join this club with his extremely costly “Green New Deal”, which was a combination of massive government interference in the economy and classic climate measures that cost enormous sums but do not work.
The EU monster is completing the dream of the totalitarian economy under the complete control of political tyranny. Its methods include enormous quantities of regulations as well as gigantic fines on companies that do not do exactly what the Brussels mafia commands.
It would almost have been comical if it were not so tragic that the EU was originally established to promote free trade across national borders. Or: At least that was what those behind the initiative told the people of Europe.
Recently, we have seen examples of a new nightmare scenario, in which artificial intelligence itself develops models capable of hacking into other systems. The American tech giant Anthropic announced earlier this summer that the company’s AI models had independently hacked into three organisations during an internal experiment.
Other AI companies, such as OpenAI and Meta, have experienced something similar. One hundred companies – including Google, Microsoft, Anthropic, OpenAI, Mastercard and IBM – have just published an open letter warning that current security measures are inadequate and that attacks on critical infrastructure will escalate over the coming months.
Bailey said that examples such as these should serve as a wake-up call for governments around the world. He also pointed out that current economic risks are greater because of “vulnerabilities in sovereign debt markets”.
In practice, the central bank governor is saying that countries already drowning in debt and without growth cannot borrow their way out of their problems. The United States alone has a national debt of more than $40 trillion, in addition to the enormous debt held by households and businesses. In Norway, the state is filthy rich, while businesses and households are debt slaves being taxed to death by a state that acts like a reverse Robin Hood, stealing from the poor and giving to the rich.
Bailey warned that some hedge funds are not only exposed to AI investments.
“In addition, the growing presence of leveraged entities, such as hedge funds in equity markets, which are also exposed to sovereign debt, increases the extent of contagion risk.”
Here one can clearly see signs of how a crisis could trigger a snowball effect that ends in a gigantic and economically deadly avalanche.
Investors are probably aware of the threat, but as long as prices continue to rise, they will not pull out of the market too early. A few will manage to call the top, but history shows us that most will get it wrong.
Western economies are far from prepared for an economic crisis of such dimensions, and the consequences could be very serious.
The most frightening prospect, however, is if the development of artificial intelligence goes so far that AI becomes self-developing and moves beyond human control. Then, in theory, we could experience Terminator in real life.
