The Bank of England has warned that rapid developments in artificial intelligence and machine learning could pose risks to the UK’s financial stability, triggering a fresh review into their use across the City.
AI and machine learning have been used by financial firms for at least a decade, for example, to help detect fraud and money laundering. However, a recent increase in technological advances and available data, as well as falling costs of computing power, had fuelled “considerable interest” and more widespread use across the sector, the Bank’s financial policy committee (FPC) said.
While most companies suggested the way they were using the technology was “relatively low risk”, the FPC warned that wider adoption “could pose system-wide financial stability risks”. That could mean greater “herding behaviour”, where there is a concentration of firms making similar financial decisions that then skews markets, and an increase in cyber risks.
The Bank’s governor, Andrew Bailey, said it was important to recognise that developments in AI had the potential to boost economic growth and productivity. However, he warned that “we have to embrace it with our eyes open”.
There are concerns that few people understand how AI works, making it harder for regulators to hold people accountable for the decisions made, or actions taken, as a result.
Bailey said it was similar to the algorithmic trading, where powerful computers analyse markets to buy or sell shares within seconds. “You can’t have people using algorithmic trading without understanding how it works. Because it’s a recipe for trouble.”
However, “It’s not out of control in terms of 2001: A Space Odyssey”, Bailey said, referring to the 1968 sci-fi film where the ship’s computer turns against its crew. “It’s actually that the thing is so complicated.
“All of us who have used it have had the experience of a sort of hallucination, and it sort of comes up with something that you think: ‘How on Earth did that come out?’
“If you’re going to use it for the real world and real financial services, you can’t have that sort of thing happening. You’ve obviously got to have controls and an understanding of how this thing works, and there’s a lot to do there.
Meanwhile, new AI and data providers could also end up becoming important participants in financial markets and may require closer oversight. The Bank, its regulatory arm the Prudential Regulation Authority, as well as the City watchdog the Financial Conduct Authority will launch a consultation paper on these “critical third parties” later this month.
The FPC said it would “consider the financial stability risks of AI and ML [machine learning] in 2024, and, working alongside other relevant authorities, would seek to ensure that the UK financial system was resilient to risks that may arise from widespread adoption”.
The deputy governor Sam Woods, who is also the chief executive of the Prudential Regulation Authority, said it was not yet clear whether the review or consultation would result in AI-specific regulation.
While the regulator had so far remained “technology agnostic”, it may explore the “steps that a senior manager must take to assure themselves that what is coming out of the black box is actually reasonable … so there are things like that that we may need to add”.
News of the fresh review was revealed alongside the Bank’s financial stability report, which suggested that mortgage repayment risks had eased.
Although mortgage holders have been particularly worried about repayments, given the rise in interest rates since 2021, the committee said households and businesses were proving to be relatively resilient in the face of higher-for-longer interest rates in the UK.
About 55% of the UK’s mortgage borrowers have so far had to reprice on to higher-rate contracts, accounting for about 5 million households. The remainder will reprice by 2026, with average monthly repayments expected to increase by £240. That is slightly down from previous expectations of an increase of £250.
While household finances remain “stretched” by higher living costs and interest rates, that has been tempered somewhat by higher-than-expected income growth, the Bank said.
It is now expecting only 1.6% of households – or 440,000 – to shell out more than 70% of their income towards essential living costs and mortgage payments in 2024. That is down from forecasts released in July, which had pointed to an uptick to 2.5%.
