FCA releases Research Note on the role of AI in credit decisions

New research from the UK’s Financial Conduct Authority (FCA) explores how different methods of explaining AI-assisted credit decisions impact consumers’ ability to identify errors. The study found that while transparency boosts confidence, too much information can sometimes make it harder for consumers to challenge mistakes.

Participants struggled to detect errors in credit scoring algorithms depending on how explanations were presented. Simply listing the data used by the AI made it harder to spot incorrect inputs but unexpectedly helped consumers challenge flawed decision logic. The findings suggest that more information is not always better and that explanation methods should be carefully tested to avoid unintended consequences.

The FCA’s AI research series aims to support the responsible use of AI in financial markets. Future research will explore how AI explanations affect consumer decision-making across different financial services.