Capital One UK research indicates that 13% of people used an AI tool for financial information or advice in the previous year. This is a clear signal that clients may arrive at mortgage and protection conversations with AI-generated assumptions about affordability, rates, credit files or product eligibility.
Advisers can turn this into an opportunity by offering a human-led fact-checking service: clarify the limits of generic AI outputs, explain lender-specific underwriting and document why a recommendation is suitable for the individual. Firms using generative AI internally should ensure outputs are reviewed for accuracy, privacy and Consumer Duty outcomes, particularly where material is client-facing.
The wider regulatory focus on fraud reinforces the point. The FCA says firms closed 238,396 suspected money-mule accounts in 2025 and stresses the need for earlier detection. AI can assist with anomaly detection and operational screening, but firms should be cautious about model bias, explainability, data quality and over-reliance on automated flags.