Each weeknight, tens of thousands of Britons open a chatbot and ask something their bank manager would once have answered: should I consolidate my pensions? Is now a good time to invest in an ISA? Could I afford to retire at 60?
The answer they receive carries no regulatory weight, no compensation backing, and no complaints pathway. Yet according to the Financial Conduct Authority’s latest Perimeter Report, published on 26 March 2026, the practice is now widespread enough that the regulator has flagged it as a systemic concern — one that existing rules were never designed to address.
The advice gap, filled by default
Fewer than one in ten UK adults currently receive regulated financial advice. Meanwhile, data from the FCA’s March report shows nearly one in five investors already rely on social media for financial decisions, and — in a striking finding highlighted by the watchdog — a growing proportion have migrated to AI tools altogether. A Sky News investigation this spring put the figure at roughly 40 per cent of British adults turning to unregulated AI guidance for at least some financial decisions, up sharply from previous surveys.
The economics are straightforward. A one-off piece of regulated financial advice in the UK now costs, on average, between £500 and £2,500 depending on complexity. ChatGPT is free at midnight. The advice gap the FCA has wrestled with for decades has not closed; it has simply moved online and gone unregulated.
What falling outside the perimeter actually means
The FCA’s position is unambiguous: general-purpose AI tools sit outside its regulatory perimeter. A consumer who acts on guidance from a chatbot has no recourse to the Financial Ombudsman Service if things go wrong. They are not covered by the Financial Services Compensation Scheme. No authorised person is accountable.
That matters when the guidance is wrong. AI models have no fiduciary duty, no continuing obligation of care, and no knowledge of a user’s full financial picture. They can confidently suggest moving savings into assets that are wholly inappropriate for a given individual’s circumstances — and no rule currently requires them to do otherwise.
The House of Commons Treasury Committee raised the alarm formally in January. Its report, published on 20 January 2026, warned that the FCA’s “wait-and-see” approach to AI in financial services “risks serious harm to consumers and the wider system”. Among the Committee’s recommendations: AI-specific stress testing of systemically significant models, clear practical guidance on how consumer protection rules apply to AI by the end of 2026, and the designation of major AI providers as critical third parties under the UK CTP regime.
The regulator’s response — and where it may move
The FCA has so far resisted AI-specific regulation, preferring a technology-neutral, principles-based approach. Its long-running Mills Review, launched on 27 January 2026, is examining how AI might reshape retail financial services by 2030 — but the watchdog has indicated it considers major regulatory changes “premature” pending that review’s findings.
There are hints of movement. FCA chief data officer Jessica Rusu told the Innovate Finance Global Summit on 21 April that the authority sees agentic commerce — where AI systems make financial decisions autonomously — as an area requiring “active regulatory shaping”. The FCA’s Simplified Advice consultation (CP26/10) is simultaneously trying to lower the cost of human advice delivery, addressing the supply side of the gap even as AI colonises the demand side.
Technology Secretary Liz Kendall reinforced the strategic stakes at a keynote on 28 April, framing AI as “central to UK economic prosperity and national security” — language that sits awkwardly alongside a regulator still watching from the sidelines.
A gap that cannot wait
What the FCA’s perimeter report makes plain is that the advice gap is no longer merely a distributional problem. It is becoming a consumer protection crisis operating in plain sight. Millions of Britons are making consequential financial decisions based on guidance that carries no accountability, no verification, and no safety net.
The Mills Review will report. The consultations will close. Guidance will eventually arrive. But each month of regulatory lag is another month in which Britons, priced out of professional advice, are navigating markets with tools the regulator cannot reach.
Discussion
Sign in to join the discussion.
No comments yet. Be the first to share your thoughts.