AI Boom, AI Bust? Why Andrew Bailey Is Warning About a Global Shock

Artificial intelligence is usually presented as an economic opportunity. It could increase productivity, automate routine work, accelerate research and create entirely new industries.

But the amount of money now flowing into AI is also creating a different kind of risk.

Andrew Bailey, Governor of the Bank of England and chair of the international Financial Stability Board, has warned G20 finance ministers and central bankers that a sharp reversal in the AI boom could contribute to a wider global economic downturn.

His concern is not that AI itself will suddenly cause a recession. The bigger issue is that financial markets are becoming increasingly exposed to the same technology companies, infrastructure providers and investment assumptions.

Why the AI boom could become a financial risk

Huge amounts of investment are now concentrated in a relatively small number of AI companies, semiconductor manufacturers, cloud providers and hyperscale technology businesses.

Many of these companies are also increasingly dependent on one another. They use the same chips, cloud infrastructure, data centres and energy supplies, while some are investing directly in each other or becoming major customers of one another.

That creates concentration risk.

If investors begin to question whether future AI revenues can justify today’s valuations, share prices could fall quickly. Where investors have borrowed heavily to increase their exposure, those falls could trigger forced selling and make the correction even worse.

What starts as a technology market correction could then spread into the wider financial system.

The risks are starting to overlap

The key concern is not one single weakness, but several risks becoming connected at the same time:

• High valuations across parts of the AI sector
• Heavy concentration in a small number of major technology companies
• Increased borrowing and leverage by investors
• Dependence on common cloud, chip and data-centre infrastructure
• Rapid growth in electricity demand from AI
• Greater exposure to AI-enabled cyberattacks
• Strong financial links between many of the same companies

Each risk may be manageable on its own. The danger increases when several happen together.

Is this another dot-com bubble?

There are obvious comparisons with the dot-com boom of the late 1990s.

Investors poured huge amounts of money into internet companies, valuations became detached from reality and many businesses eventually failed. Yet the internet itself still went on to transform the global economy.

AI could follow a similar pattern.

The technology may prove genuinely revolutionary while many of today’s investment assumptions still turn out to be wrong.

The difference is that today’s largest technology companies are much more deeply embedded in the global economy than many dot-com businesses were. They provide cloud services, computing infrastructure and digital systems used by banks, governments and major corporations.

That potentially makes any major disruption more widespread.

Cyber risk is part of the same picture

Bailey has also highlighted the growing cyber threat created by advanced AI.

AI can help organisations identify vulnerabilities and defend their systems, but the same technology can also help attackers automate cyberattacks, write malicious software and identify weaknesses much faster.

The financial system is particularly exposed because many organisations depend on the same technology providers.

A serious failure or cyberattack affecting a major cloud or software platform could therefore disrupt many banks and businesses at the same time.

This is very different from the traditional assumption that one organisation fails while everyone else continues operating normally.

AI could still be good for the economy

None of this means AI is necessarily bad for economic growth.

Bailey has also argued that AI and robotics could become major general-purpose technologies, potentially improving productivity across many industries.

The danger is more about expectations.

Investors may be pricing in enormous future profits long before those profits actually appear. If reality fails to match those expectations, markets can fall sharply even while the underlying technology continues to improve.

That has happened before with railways, telecommunications and the internet.

The Lomond Logic view

The most important part of Bailey’s warning is not simply the possibility of an AI bubble.

It is the convergence of risk.

Financial markets, technology infrastructure, energy demand, cyber security and corporate concentration are increasingly connected. A shock in one area can therefore travel quickly into others.

A fall in AI valuations could reduce investor confidence, trigger forced selling, tighten credit conditions and slow investment. At the same time, dependence on shared technology infrastructure could amplify the impact of a major operational failure or cyberattack.

That is what turns AI from a technology story into a systemic-risk story.

The AI revolution may still deliver enormous economic benefits. The real question is whether financial markets and critical infrastructure are becoming too dependent on the assumption that everything goes right.