AI in retail investing: the FCA's warning shot on unregulated advice

The FCA has quantified what advisers and platforms have suspected for a year: AI has moved from novelty to primary research tool for a generation of retail investors. Four in five less experienced investors aged 18 to 40 have used AI for help with investing, and 56% trust AI tools more than TV and radio (47%), press (46%) or social media influencers (29%) (FCA). The regulator's concern is not adoption. It is the confusion about protection that comes with it.

Key Executive Takeaways

• The FCA has found that 44% of young investors wrongly believe AI-generated financial information is regulated, and 32% incorrectly think FSCS or Financial Ombudsman Service cover would apply if AI advice went wrong (FCA).
• General purpose AI chatbots sit outside FCA authorisation, but tools specifically deployed to give financial advice would likely fall within its remit, creating a live perimeter question for any firm embedding third-party models in customer journeys (FCA).
• With 38% of respondents willing to invest on AI output alone and two-thirds expecting to rely on it more over the next year, boards should treat AI-mediated decisions as a Consumer Duty risk vector, not a marketing opportunity (FCA).

A perimeter problem, not a technology problem

The substantive shift in the FCA's framing is jurisdictional. Lucy Castledine, director of consumer investments at the FCA, said AI 'can help you research companies, understand jargon or explore options before you make a decision' but that consumers 'need to understand how you're protected and continue to use your own judgement' (FCA). That is a careful line. The regulator is signalling that a tool 'specifically set up to provide financial advice would be likely to fall within the FCA's remit', which puts pressure on any firm procuring or fine-tuning models that respond to product-specific queries.

For platforms, the question is where an educational chatbot ends and personal recommendation begins. Firms have historically managed this boundary with disclaimers and content controls. Generative interfaces make that harder: the same model that summarises a fund factsheet can, with a slightly different prompt, produce something that reads like a recommendation. If a third of users already think FSCS applies, the reputational tail on a bad outcome is longer than the regulatory one.

Consumer Duty implications for distribution

The research also lands in the middle of the FCA's broader Consumer Duty enforcement posture. The regulator is already scrutinising fair value and vulnerable customer access in the Child Trust Fund review reporting next year (FCA). Applied to AI, the same fair value logic asks whether a firm's target market assessment reflects how customers actually reach investment decisions. If most under-40s are triangulating with a chatbot before opening a stocks and shares ISA, distribution strategies and financial promotion controls that assume a linear funnel are out of date.

There is also a supervisory information asymmetry. Firms can see their own funnels; they cannot see what an unregulated chatbot told the customer in the fifteen minutes before they arrived. That gap makes complaints handling, suitability records and appropriateness testing harder to defend.

What to do before the FCA does it for you

Senior leaders should commission two pieces of work now: a perimeter review of any AI features that touch investment content, tested against the FCA's remit signal; and a Consumer Duty gap analysis that assumes customers are arriving pre-influenced by unregulated AI. The firms that treat this as a compliance exercise will be caught by the ones that treat it as a distribution redesign.