Britain Needs 'AI Stress Tests' for Financial Services, Lawmakers Say

A cross-party group of UK lawmakers has called for AI-specific stress tests for financial services to mitigate the risks posed by the increasing use of artificial intelligence in the sector. The Treasury Committee criticized the current approach of the Financial Conduct Authority (FCA) and the Bank of England as a 'wait and see' stance, which it said is insufficient to manage AI-related risks to consumers and financial stability. The report emphasized the need for proactive measures to address potential issues arising from AI systems.

The committee warned that AI-driven trading could amplify herding behavior, potentially leading to market instability or even a financial crisis in a worst-case scenario. About three-quarters of UK financial firms now use AI in core functions like insurance claims processing and credit assessments. While these applications bring benefits, the technology also introduces significant risks such as algorithmic bias and unregulated financial advice from chatbots.

Experts contributing to the report highlighted the risks of over-reliance on U.S. tech giants for AI and cloud services, which could compromise the resilience of the UK financial system. The Treasury Committee recommended that the FCA publish detailed guidance by the end of 2026 on consumer protection rules in relation to AI. It also urged the government to designate AI and cloud providers as critical third parties to improve regulatory oversight.

Why Did This Happen?

The committee chair, Meg Hillier, expressed concern that the UK financial system is unprepared for a major AI-related incident. The report cited evidence showing that many financial professionals do not fully understand the risks posed by AI systems. This includes the risk of AI-driven decisions amplifying existing vulnerabilities and creating a false sense of security.

What Are Analysts Watching Next?

The report also emphasized the need for stress tests to assess the resilience of financial institutions in the face of AI-driven market shocks. These tests would help regulators identify potential weaknesses and ensure that the financial system is prepared for unexpected events. The Treasury Committee called on the FCA and the Bank of England to collaborate on this effort to strengthen the UK's financial infrastructure.

The report also outlined the risks associated with AI in areas like cybersecurity and data privacy. Financial firms are becoming increasingly reliant on AI to automate administrative tasks and deliver core services. However, this automation introduces new vulnerabilities, especially when AI systems make decisions without clear human oversight.

What Are the Next Steps for Regulators?

The committee recommended that the FCA take a more active role in clarifying accountability for AI-related harms. It called for clear guidance on who should be responsible when AI systems cause issues for consumers. The report also noted the importance of ensuring that senior managers understand the systems they oversee.

The Treasury has taken some steps to address AI risks, including the appointment of two new 'AI Champions' from Starling Bank and Lloyds Banking Group. These individuals will help steer the safe adoption of AI in financial services. The government aims to strike a balance between managing risks and unlocking the potential of AI for economic growth.

The FCA has stated it will review the report and consider the recommendations. The Bank of England did not respond to a request for comment. Both institutions have previously expressed concerns about the fast pace of technological change and the challenges of creating specific AI rules.

The report highlights the urgent need for regulatory action to ensure that the UK financial system is prepared for the risks posed by AI. With more than three-quarters of financial services firms already using AI, the implications for consumer protection and financial stability are significant. The Treasury Committee emphasized that proactive oversight is essential to prevent serious harm to the public and the economy.