The Financial Conduct Authority published research on 26-27 August reporting that young British investors trust AI tools more than they trust television, the press or social-media influencers. The headline is doing the least useful work in the release.

The trust ranking

Among UK adults aged 18 to 40 who own or are considering investments, 56% said they trust AI tools. Television and radio scored 47%, the press 46%, and social-media influencers 29%. As a ranking it is unsurprising: a tool that answers your specific question in your own words will out-poll a broadcast medium that answers nobody's, and influencer trust has been falling for years independently of AI.

The finding underneath

Two other numbers do more work. 44% mistakenly believe AI-generated financial information is regulated. 38% consider it acceptable to make an investment decision on AI output alone. And 32% wrongly believe they would be entitled to compensation from the Financial Services Compensation Scheme or the Financial Ombudsman Service if AI-driven advice lost them money. That last one is a concrete, testable false belief about a statutory protection, held by roughly a third of the sample, and it is the one with a direct consumer-harm pathway: someone acts on a chatbot's suggestion, loses money, and discovers no redress exists.

What the common framing gets wrong

Coverage is running the trust comparison as though it establishes a shift in where a generation gets financial information. The base does not support that weight. This is 666 respondents, recruited through the online panel platform Attest, fielded on a single day — 24 July 2026. An online panel of people who already use the internet enough to join one will over-represent AI familiarity relative to the general population by construction. As a directional indicator of a misconception the FCA wants to correct, it is fit for purpose. As evidence that a generation has switched information sources, it is a small single-day convenience sample being asked to carry a demographic conclusion.

Why the regulator published it anyway

The FCA's interest is not the ranking. A regulator that finds a third of a surveyed group believes a statutory compensation scheme covers something it does not has identified a communications problem it is obliged to act on, and the fastest way to act is to publish the finding so the correction reaches people through the press. Read that way, this is a consumer warning wearing a survey's clothes — and the sentence worth carrying out of it is that neither the FSCS nor the Ombudsman covers losses from following a chatbot.