Industry commentary has highlighted the risks of shadow AI, where staff use unapproved tools to process client information or generate communications outside an established control framework. At the same time, the FCA continues to emphasise strong compliance cultures and reports that firms’ understanding of Consumer Duty expectations has improved.
Mortgage and financial-advice firms deploying AI should define approved use cases, restrict sensitive data sharing, test outputs for accuracy and bias, and retain meaningful human oversight. A practical starting point is to use AI for low-risk administrative tasks while maintaining formal review and record-keeping for client-facing content, advice support and compliance monitoring.