Artificial intelligence has become one of the biggest investment stories in the global economy. But according to Bank of England Governor Andrew Bailey, the speed and scale of the AI boom could also create new risks for financial stability.
Bailey, who also chairs the Financial Stability Board, has warned G20 finance ministers and central bank governors that a sharp correction in AI-related markets could have consequences far beyond the technology sector.
The warning comes as investors continue to place enormous expectations on companies developing AI infrastructure and technology. Nvidia, one of the most important companies in the AI supply chain, recently reached a market valuation of around $5.1 trillion, highlighting the extraordinary amount of capital flowing into the sector.
The concern isn’t AI itself
Bailey’s warning is not that artificial intelligence will inevitably cause an economic downturn. Instead, the concern centres on what could happen if expectations surrounding AI become disconnected from the technology’s ability to generate sustainable economic returns.
AI-related companies have attracted significant investment, while equity valuations have risen sharply. The Bank of England’s July Financial Stability Report noted that AI-related stocks had become an increasingly important part of global equity markets, with valuations becoming more stretched and market concentration increasing.
If investors suddenly reassess the expected value of AI companies, markets could experience a rapid correction.
That could become more serious if large amounts of debt are involved.
An AI investment boom increasingly backed by debt
Developing and deploying advanced AI requires enormous amounts of infrastructure, including data centres, computing capacity, semiconductors and energy.
The Bank of England says AI companies have increasingly turned to external financing, including private credit, public debt markets and bank lending, to fund this expansion. It described the pace of investment as historically unprecedented.
This creates a potential chain reaction.
If expectations for future AI growth decline, companies could face lower valuations and weaker financing conditions. Highly leveraged businesses could then find it harder to service their debt, potentially affecting lenders and investors.
The concern becomes greater because AI-related companies are increasingly interconnected with the wider financial system.
Cybersecurity could be an even bigger risk
Market valuations are only one part of Bailey’s warning.
Advanced “frontier” AI systems are also creating new cybersecurity and operational risks. The Bank of England has warned that increasingly capable AI models could make cyberattacks faster, easier to execute and potentially more disruptive.
For the financial sector, this matters enormously.
Banks, payment providers, investment firms and other financial institutions depend on interconnected digital infrastructure. A sophisticated AI-enabled cyberattack affecting multiple organisations could potentially create disruption across markets rather than remaining isolated to a single company.
The Financial Stability Board has identified AI-driven cyber risk as a particularly pressing concern for global financial stability.
The paradox of AI
There is an important contradiction at the heart of the debate.
AI could become one of the most powerful productivity technologies in modern economic history. It could help businesses automate repetitive work, improve decision-making, increase efficiency and create new products and industries.
The Bank of England itself acknowledges that AI could generate significant productivity gains and support long-term economic growth.
But the economic benefits are still uncertain.
Markets are currently pricing in expectations about how quickly AI adoption will spread, how much productivity it will generate and how successfully companies will monetise it. If those expectations prove too optimistic, financial markets could react before the underlying economic benefits have time to materialise.
That creates a difficult balancing act for policymakers.
What should happen next?
Bailey has called for stronger international coordination around the testing and deployment of frontier AI systems. His argument is that AI risks cannot be addressed effectively by individual countries because technology, financial markets and cyber infrastructure operate across borders.
For financial institutions, resilience will also become increasingly important.
Banks and other critical organisations need to be prepared not only to prevent cyberattacks but also to recover quickly if their systems are compromised. Regulators are therefore facing the challenge of keeping pace with AI development while ensuring that financial institutions remain resilient.
The bigger lesson
The AI boom does not have to end in an economic crisis.
The bigger lesson from Bailey’s warning is that technological progress and financial stability need to develop together.
AI may ultimately deliver enormous economic benefits. But when investment grows extremely quickly, valuations become concentrated and increasingly sophisticated technology introduces new operational risks, the consequences of a sudden loss of confidence can become much larger.
The challenge for governments, regulators, investors and businesses is therefore not to stop AI innovation.
It is to make sure the financial and technological systems surrounding it are resilient enough to handle both its success — and the possibility that expectations sometimes run ahead of reality.
AI may be one of the defining technologies of the next economic era. The question is whether the world can manage its financial risks as quickly as it develops its capabilities.
