BSA response to the FCA's Mills Review into the long-term impact of AI on retail financial services. 

The BSA polled members on the four themes set out in the Mills Review, with an additional poll on security and protections.

We also asked about current personal and organisational usage of AI and received equal responses from sm all £50m--£500m in assets), medium (£500m-£1.5bn in assets), and large (£1.5bn+ in assets) mutuals.

In terms of current usage of AI, this varies across the sector. Most organisations across all size categories are conducting small pilots or team-level trials. A small number of organisations have progressed to moderate or significant use.

The BSA's full response and answers to the FCA's questions on market adoption and future regulatory approach can be found here. 

Summary of views

1. Future Evolution of AI: Members expect AI to evolve as a tool for internal operational efficiency and enhanced risk management. The focus is on automating administrative workflows, such as case packaging and document verification, while strengthening fraud detection and prevention via anomaly spotting.

2. Future Impact on Markets and Firms: While there is significant uncertainty regarding market concentration, most members believe traditional financial institutions will retain their primary role as the customer interface. However, there is a recognized threat from technology platforms that could leverage large ecosystems to disrupt this relationship.

3. Future Consumer Trends: The technology is expected to make financial information more accessible and easier to understand, potentially improving financial literacy. Despite these benefits, members identify a growing risk of consumer vulnerability driven by over-reliance on automated advice and a lack of scrutiny regarding AI accuracy.

4. Future Regulatory Approach: Most of the responses we received view existing frameworks as only partially fit for purpose. From a governance perspective, members prefer a regulatory approach that focuses on fairness, accuracy, and transparency, rather than a prescriptive focus on the underlying technology itself. Among the responses we received, there is strong support for independent third-party testing and certification of AI systems. They are also looking for government-backed assurance and clearer guidance from the FCA on acceptable and unacceptable practices would significantly bolster their risk management confidence.