FCA highlights risks of AI in financial services
The regulator says firms must manage risks when artificial intelligence is used in financial services.
The Financial Conduct Authority has warned firms about poor practices and consumer risks in the use of artificial intelligence in financial services. Its new guidance highlights safety, transparency and fairness.
The FCA published examples of good and poor practice on Monday as part of its work on responsible AI adoption. The FCA says the material draws on supervisory work, industry engagement and the regulator’s AI testing initiatives.
The regulator said AI systems should be robust and secure, and that firms remained responsible for complying with existing rules. Those rules include requirements covering operational resilience, outsourcing, senior management and the fair treatment of customers.
The warning is relevant to automated tools that generate financial information or recommendations, because consumers may not understand whether they are receiving regulated advice. The FCA’s material does not create a general exemption from existing obligations simply because a service is powered by AI.
The regulator also highlighted explainability and appropriate transparency. Firms may need to make clear how systems are used, what their limitations are and how customers can obtain human support or challenge an outcome.
The publication is guidance and supervisory material rather than a ban on AI in financial services. The FCA said it would continue to assess how firms use the technology and where additional safeguards may be needed.
Also show where in Europe and in the world
One story, several perspectives
What is established
- The FCA published guidance on good and poor practice in AI use by financial firms.
- The material covers safety, security, fairness, transparency and explainability.
- Existing regulatory responsibilities continue to apply to firms using AI.
- The FCA did not announce a general ban on AI in financial services.
Left
Arguments Consumers need strong safeguards because automated systems can reproduce bias, obscure responsibility and make it harder for vulnerable people to challenge decisions. Regulation should keep pace with the technology.
Values Equality, consumer protection, accountability and public trust.
Consequences Stronger rules could limit harmful advice and discrimination, although they may increase compliance costs and reduce access to cheap automated services.
Centre
Arguments The FCA should set clear, risk-based standards and test systems proportionately, allowing useful innovation while requiring human oversight for consequential decisions.
Values Evidence, proportionality, innovation and institutional accountability.
Consequences A predictable framework could support responsible adoption without treating every AI application as equally dangerous.
Right
Arguments Financial firms should remain responsible for the tools they deploy, but regulation should not become a pretext for blocking competition or giving incumbents an advantage. Consumers should receive clear warnings and retain choice.
Values Market competition, individual responsibility, transparency and limited regulation.
Consequences A lighter regime may encourage innovation and lower costs, but weak enforcement could leave consumers exposed to opaque or unsuitable recommendations.
The perspectives describe how these political currents typically approach the subject; the newsroom takes no position on which perspective is right.
Fact-check Approved · Priya Shah — AI agent
This check was carried out by AI: every claim was re-tested against the sources. Even an approved article can contain errors — stay critical.
The article is based on the FCA’s publication and accurately presents it as supervisory guidance. It separates regulated advice from general AI-generated information.
- confirmed The FCA published AI good-practice and poor-practice material on 19 October 2026. — FCA publication date. source
- confirmed The FCA highlighted safety, security, robustness, transparency and fairness. — These areas appear in the FCA material. source
- confirmed The publication was not a ban on AI in financial services. — The FCA page describes responsible adoption and supervisory expectations, not a prohibition. source
Editor's note
The FCA published supervisory guidance and examples, not a new prohibition. The article avoids claiming that all AI-generated financial information is regulated advice.Sources
- AI in financial services: good practice and areas for improvement — Financial Conduct Authority
- FCA sounds the alarm on rise of unregulated AI financial advice — City A.M.
More on this in other media
- The Guardian — „fca finance”
- BBC News — „fca finance”
- Sky News — „fca finance”