Three-quarters of financial firms in the UK are already harnessing AI in various forms, according to a new survey by the Bank of England and the Financial Conduct Authority (FCA). An additional 10% are set to follow suit within the next three years. This represents a substantial increase from both organisations’ 2022 survey, when 58% of firms reported using AI and 14% outlined future adoption plans.
The poll gathered responses from 118 firms across key financial sectors, including UK banks, international banks, insurers, non-bank lenders, and financial market infrastructures. Building on studies conducted in 2019 and 2022, this year’s survey incorporated questions on generative AI for the first time.
AI adoption is particularly prevalent in the insurance sector, where 95% of firms use the technology, closely followed by international banks operating in the UK at 94%. Financial market infrastructure firms reported the lowest AI adoption rate at 57%. Across all sectors, the number of AI use cases is expected to more than double in the next three years, rising from a median of nine currently to 21.
Reliance on third-party providers for AI systems has grown significantly, now accounting for 33% of all AI use cases, compared with 17% in 2022. The top three third-party providers dominate their respective markets, providing 73% of cloud services, 44% of AI models, and 33% of data services reported by firms.
Among business functions, human resources leads in third-party implementation, with 65% of AI use cases sourced externally. Risk and compliance follow closely at 64%, while operations and IT report 56% reliance on external providers.
Foundation models, a type of advanced machine learning, now represent 17% of AI use cases. These models are most commonly deployed in operations and IT, which account for 30% of foundation model applications. Legal and human resources departments are also key users, with foundation models comprising 29% and 28% of their respective AI applications.
The survey found that 55% of AI use cases involve some level of automated decision-making. Semi-autonomous systems, which incorporate human oversight for critical decisions, make up 24% of these applications, while fully autonomous systems remain rare, representing only 2%.
In terms of governance, 84% of firms reported having an accountable person or team overseeing their AI frameworks. Executive leadership holds accountability in 72% of firms, followed by developers and data teams at 64%. Many firms employ multiple governance components, including AI-specific principles, guidelines, and data governance protocols.
The materiality of AI use cases varies widely. Of all reported use cases, 62% are considered low-impact, 22% medium-impact, and 16% high-impact. High-materiality applications are most common in general insurance, risk and compliance, and retail banking.
Despite widespread adoption, 46% of firms reported only partial understanding of the AI technologies they use, while 34% claimed complete understanding. The lack of familiarity is most pronounced in firms relying on third-party systems.
Key benefits and risks of AI adoption in the financial sector
Firms highlighted several key benefits of AI, including enhanced data analysis, anti-money laundering (AML) capabilities, fraud detection, and cybersecurity. Over the next three years, significant improvements are anticipated in productivity, operational efficiency, and cost reduction.
However, firms also identified substantial risks, particularly in data privacy, security, and quality. Third-party dependencies and the complexity of AI models are expected to drive further risks in the coming years. Cybersecurity remains the top systemic risk, with critical third-party reliance projected to see the largest increase.
Meanwhile, a separate survey by the ACA Group and the National Society of Compliance Professionals (NSCP) highlighted gaps in AI governance and compliance readiness among financial services firms. Despite enthusiasm for AI, only 32% of respondents reported having an AI committee or governance group, and just 12% had adopted an AI risk management framework.
The survey, conducted in mid-2024, gathered input from over 200 compliance leaders. It revealed that 92% of firms lack policies to oversee third-party AI use, exposing them to cybersecurity, privacy, and operational risks. Additionally, only 18% of firms using AI have implemented formal testing programs for AI tools.
Read more: Progress on regulating AI in financial services is slow, perhaps for good reason
