AI in financial services: the FCA's good and poor practice adds no new rules
The FCA's review of AI in retail financial services sets out good and poor practice in three areas and says plainly that it does not create new requirements. The Consumer Duty and SM&CR carry the weight.
What was published
On 19 October 2026 the FCA published good and poor practice on AI in financial services in three chapters: Consumer Duty and vulnerability; governance and accountability; and operational resilience. It states that the material does “not create new requirements” 48. It follows the Mills Review of 6 July 2026, which made seven recommendations 47.
What to do
- List the decisions in your customer journey where a model, scorecard or AI tool makes or shapes the outcome.
- For each, show how your outcomes monitoring under PRIN 2A.9.9R would detect a group of customers doing worse because of it.
- Name the senior manager who owns each one, and record the oversight in the evidence your governing body reviews under PRIN 2A.8.4R.
Our view
No new rules means no new excuses. The questions a supervisor will ask about an AI-driven decision are the questions the Duty already asks of any decision: who it affects, how you know the outcome is good, and who is accountable. Firms that can answer those for their human processes can extend the same file to their models.
How this was produced: drafted from the source and our Handbook mirror, and reviewed by Automated Compliance before publication.