New data from the Financial Conduct Authority has brought the FCA young investor AI trust question into sharp focus, finding that almost half of young investors wrongly believe AI-generated financial information is subject to regulatory oversight. The regulator commissioned the platform Attest to survey 666 UK adults aged 18 to 40 who own or are considering investments, with fieldwork completed on 24 July 2026, according to Fintechly.
The findings present a picture that any long-term portfolio manager should take seriously. Young investors are not simply browsing AI tools for curiosity; a meaningful share are treating AI output as they might treat regulated guidance, without the protections that regulated guidance actually carries.
Where AI now sits in the FCA young investor AI trust hierarchy
The survey found that 56% of respondents trust AI tools for financial decision-making, placing artificial intelligence ahead of TV and radio at 47%, the press at 46%, and social media influencers at 29%. That AI should outrank the press and broadcast media among this cohort is itself instructive. What it tells the FCA, and should tell portfolio-aware savers, is that the information environment for younger investors has shifted in ways that existing regulatory frameworks were not designed to address.
The regulator’s concern is not that AI tools exist, but that the assumed protections around them do not. A regulated financial adviser operates under FCA conduct rules, must assess suitability, and carries professional indemnity obligations. An AI chatbot, however sophisticated its output, carries none of those. When a young investor acts on AI-generated information in the belief that it is regulated, the gap between perception and reality can translate directly into portfolio risk.
Portfolio implications for those building long-term wealth
For someone in the accumulation phase, perhaps managing a self-invested personal pension (SIPP) over a ten-to-twenty-year horizon, the practical danger is not that AI gives obviously bad answers. It is that AI can give plausible, confident-sounding answers that are poorly suited to an individual’s tax position, risk tolerance or time horizon, and that those answers carry no accountability if they prove wrong.
As Kenneth Lamont, principal in manager research at Morningstar, observed: ‘Retail investors have a poor record of timing markets. Combine that with hyped investment narratives, volatile stocks and leverage, and mistakes can be magnified very quickly.’ AI tools, especially those optimised for engagement rather than caution, are capable of generating precisely the kind of hyped narratives Lamont describes, without any obligation to flag sequence-of-returns risk or the tax consequences of drawdown decisions.
That is not an argument against technology. AI can be genuinely useful for filtering information, comparing product types or understanding broad asset-class characteristics. The risk lies in treating a capable information retrieval tool as though it were a chartered financial planner with fiduciary responsibility. One is regulated; the other is not.
For investors approaching or already in retirement, the stakes are different but no less real. Capital preservation and income reliability depend on advice that accounts for individual circumstances. Personalisation is precisely what an unregulated AI tool cannot provide, regardless of how authoritative its tone.
The FCA’s survey covered adults up to the age of 40, so the cohort in question has time on its side. But habits formed early in an investor’s lifecycle tend to persist. If a generation learns to treat AI output as regulated guidance, that misunderstanding does not automatically correct itself as portfolios grow and the consequences of poor decisions become harder to absorb. The FCA’s data suggests the regulator is aware of that risk; whether its forthcoming guidance will be sufficient to change investor behaviour is a separate question.
