What BoE AI Risk Warnings Mean for the Financial Sector

Traditional financial institutions have historically approached disruptive technological innovations with a degree of healthy scepticism, prioritising systemic safety over rapid adoption. 

This cautious stance aligns with a warning from the Bank of England (BoE) where it warns that AI poses a growing threat to financial stability as investors bet heavily that the technology will prove a success.

However, this simultaneously increases banks’ vulnerability to cyberattacks.

In a half-yearly assessment of risks to Britain’s financial system, the central bank says previous dangers it had identified have not gone away.

These include stretched share price valuations, high public debt and risky private credit lending to businesses.

However, the BoE highlights additional dangers that have escalated since its last review.

These include a potential stock market bubble, heightened cybersecurity vulnerabilities and AI companies’ increasingly complex and opaque debt.

The central bank notes that the likelihood of multiple vulnerabilities crystallising at the same time has increased.

According to the report, this overlapping pressure is potentially amplifying its combined impacts on financial stability.

Market valuations and the risk of share corrections

For investors’ bets on AI to pay off, the BoE says there must be widespread profitable adoption of the technology. This requires the effective build-out of new infrastructure and easy access to finance for the sector.

The valuations of AI companies are currently based on earnings forecasts, which are highly uncertain.

Some of the world’s most valuable companies, like chipmaker NVIDIA, have seen their share prices soar as they invest in AI or benefit from market demand.

Those valuations have also become more stretched with the market rise partly driven by a narrow set of tech firms. The biannual report notes this is increasing market concentration in some global indices.

A sudden shift in investor sentiment regarding these businesses could trigger a sharp market correction.

A reassessment of these prospects could trigger a fall in equity prices that might be amplified by high concentration, correlated momentum-driven positions that can exacerbate volatility as markets fall, and increased leverage, the BoE says.

The Financial Policy Committee (FPC) questioned the basis for the anticipated economic benefits of using the technology, noting uncertainty over the scale and timing of future productivity gains. It remains unclear whether companies can consistently make money from AI applications.

However, the FPC adds that AI has the potential to increase productivity across a range of sectors.

The technology could still support long-term economic growth, having already supported growth in some regions.

Opaque debt structures threaten financial resilience

Corporate borrowing by tech infrastructure providers has increased rapidly, leading to a pace of investment that is unprecedented historically. 

The BoE highlights additional dangers from investors, including hedge funds, borrowing to buy shares alongside AI-related companies borrowing heavily to fund investments.

This rapid growth in leverage could pose a risk to financial stability due to increasing complexity and opacity in debt structures.

A lack of transparency about how these entities borrow could worsen a financial crisis.

Considerations around the future earnings potential for AI-related companies will also be relevant to the sustainability of these companies’ debt, the BoE adds.

If future earnings fall short of expectations, servicing this debt could become unsustainable.

The FPC exists to ensure the UK financial system can handle economic shocks. It concluded the financial system has remained resilient and continues to support the UK economy.

Cyber threats escalate via cutting edge models

Regulators globally have begun to focus more keenly on the operational impact of advanced technologies. These range from risks associated with frontier AI models such as Anthropic’s Mythos to the challenges posed by autonomous agentic systems.

Advanced AI models are increasingly capable of launching cyberattacks at greater scale.  This is due to rapid progress in cutting-edge AI since the last financial stability report was published in December.

This technological progress presents a significant increase in the risks to financial stability. It requires firms and authorities to revisit whether the resilience of key technology providers is sufficient.

In the report, the BoE says it is unclear if better AI strengthens the hand of attackers or those seeking to defend financial systems. 

However, it is likely to require more frequent software updates by financial firms, which themselves carry a risk of operational disruption.

Urging bespoke AI regulation frameworks

The rise of autonomous technologies has prompted senior central bankers to call for entirely new regulatory frameworks. 

Sarah Breeden, Bank of England Deputy Governor, signalled the need for bespoke AI regulation to contain risks posed by increasingly capable agentic systems.

The traditional regulatory toolkit was not designed to monitor autonomous digital workflows that operate without human oversight.

Sarah Breeden explains that the current systems are poorly equipped to manage these independent tools.

“Our frameworks were not built to contemplate autonomous agents, and relying on a human in the loop for all agent actions is unlikely to be realistic,” she says.

As financial organisations integrate autonomous systems, establishing these tailored frameworks will be critical to mitigating the overlapping risks detailed by the central bank.