Featured in Investment Week:
Artificial Intelligence adoption by Financial Services Firms is growing rapidly despite high levels of uncertainty about how it works, an industry survey has found.
According to research conducted by the Bank of England and the Financial Conduct Authority on AI in UK Financial Services, three quarters of firms are already using AI, with scope for a further 10% to implement the technology over the next three years.
This marks a significant uptick in the figures from the 2022 joint BoE and FCA Machine Learning in UK Financial Services survey, where the responses stood at 58% and 14%, respectively.
But despite the rush to implement this technology, almost half (46%) of respondent firms admitted to only having ‘partial understanding’ of the AI technologies they used, compared to the 34% that claimed they have a ‘complete understanding’.
According to the BoE and FCA, this is largely due to the use of third-party models, where respondent firms noted a “lack of complete understanding ompared to models developed internally”.
In 2022, 17% of AI use cases by companies surveyed were third-party implementations, but this figure has almost doubled since. The survey also found that aspects of Financial Regulation were deemed to provide constraints to firms using AI.
Data protection and privacy came out on top as the largest constraint on AI implementation, followed by resilience, cybersecurity and third-party rules and then the FCA’s Consumer Duty Regulations.
Firms also cited safety, security and the robustness of AI models as the largest non-Regulatory constraint to progress, followed by insufficient talent and a lack of access to skills. Additionally, 84% of firms said they have an accountable person for their AI framework, leaving 16% without someone spearheading the technology.
Cédric Cajet, Product Director at NeoXam, said: “As the findings suggest, firms are fast realising the tremendous benefits of AI, which can be harnessed to expedite coding and software development tasks, review Investment Data and even draft Investment Reports.
“And yet, that firms perceive Data Quality as one of the top five Risks of this new technology is hardly surprising, as many financial institutions lack a central view of their underlying Data.”
As a result, Cédric Cajet argued: “Firms must first ensure they can normalise, validate and consolidate the full breadth of Data they interact with before rushing to integrate AI. There is little point in arming yourself with the very best AI weaponry only to see it backfire”.
Meanwhile, Marion Leslie, Head of Financial Information at SIX, noted the BoE and FCA survey “identifies operational efficiency as the area with the largest expected increase in benefits.”
She added: “With the global drive for increased transparency over the past decade resulting in more Regulatory Reporting requirements for all financial institutions, it is possible they are recognizing the potential that AI holds to support this.”
