Artificial intelligence has moved from the margins of financial services to the centre of the regulatory conversation. The Bank of England’s Financial Policy Committee has warned that recent rapid advances in frontier AI capabilities have increased financial stability risks, with particular attention on cyber security and the operational resilience of the institutions that keep the financial system running.
The warning matters because AI is no longer a peripheral tool. Banks, insurers and asset managers now use machine learning across trading, credit decisions, fraud detection and customer service, and the models involved are becoming steadily more powerful and more interconnected.
Where the risks concentrate
Three concerns dominate the regulatory discussion. The first is concentration: if many institutions draw on a small number of AI providers and models, a flaw in one system can ripple through the sector. The second is operational resilience, as institutions increasingly depend on automated processes whose behaviour is harder to inspect. The third is cyber security, where the same capabilities that help detect attacks can also make them more sophisticated.
Regulators are not calling for a slowdown in adoption. The Bank’s position, set out through its financial stability reporting, is that the technology brings real benefits alongside real risks, and that the task is to manage the transition rather than resist it.
Institutions themselves are adjusting. Larger banks have begun extending their model risk management frameworks to cover machine learning systems, stress-testing the failure modes of automated processes in the same way they have long tested credit and market risk.
New standards for a fast-moving technology
Global regulators have been moving in the same direction. The Financial Stability Board has finalised a set of sound practices for the responsible adoption of AI in financial services, identifying twelve practices that help institutions harness the technology’s benefits while managing its risks. The framework covers governance, risk management and disclosure, giving firms a common reference point as usage deepens.
International bodies have echoed the theme. The International Monetary Fund has framed its latest financial stability work around resilience, risk and transformation, a formulation that captures the balance institutions are being asked to strike between innovation and caution.
What it means for the City
For UK financial institutions, the direction of travel is clear. Firms that adopt AI without strengthening their governance around it can expect growing scrutiny, while those that build resilience into adoption plans are likely to find regulators supportive.
The prize remains considerable. Institutions that manage AI risk well stand to gain real advantages in efficiency, fraud prevention and customer service. The Bank’s message is not that the technology should be feared, but that its risks deserve the same seriousness as its promise.
There is also a competitive dimension. London’s position as a global financial centre has historically rested on its willingness to adopt new technology early and govern it well, and the current moment tests both instincts at once.
For an industry that has absorbed wave after wave of technological change, that is a familiar position: the tools change, the discipline does not.