Defaqto research finds most people can’t judge whether AI guidance is accurate, with nearly one in five already acting on advice that proved wrong
Eight in ten UK consumers now use an AI assistant at least three times a week, and six in ten turn to AI to help them make decisions, according to research presented at Defaqto’s 2026 Data of Record conference on 16 September. More people are now using AI-powered tools as a source of guidance when choosing financial products than are going directly to a provider or speaking to a financial adviser, the research found.
The findings, based on consumer research carried out by Savanta on Defaqto’s behalf, point to a trust gap that has failed to keep pace with adoption. One in five consumers said they actively distrust AI tools when selecting a financial product, while almost one in five said they have acted, or nearly acted, on AI guidance that later turned out to be wrong. Most strikingly, 62% of consumers said they do not have enough information to judge whether AI-generated advice or guidance is accurate.
Mike Piddock, interim chief executive of Defaqto, told delegates that greater reliance on AI increases rather than reduces the importance of data quality, with firms needing confidence in where information has come from, how it has been structured and independently checked, and whether it offers sufficiently broad market coverage. Alex Ward-Booth, director at Savanta, likened the challenge to the California Gold Rush, where success depended on being able to distinguish genuine gold from “fool’s gold.”
Keynote speaker Emma Boardwell, founder of Emotional Finance, told the conference that firms would need to “zig towards speed, efficiency, automation and AI in order to compete” while also “zag towards human connection in order to differentiate and matter.”
For UK wealth managers and advisers, the findings sharpen a question that has been building for some time: a client who cannot evaluate the accuracy of AI-generated guidance is also poorly placed to self-correct when that guidance proves wrong, and a growing share of the market may now be making significant financial decisions from exactly that position. That reinforces the case for advisers to be explicit with clients about what independently verified, regulated advice actually provides that a generic AI assistant cannot: a named professional who is accountable for the recommendation, a properly researched view of the whole market, and someone to call when things go wrong.
Defaqto used the conference to preview changes to its Star Ratings methodology for 2027 and outlined plans to raise consumer awareness of the independent research underpinning them, framing the wider theme of the event as one where the growth of AI is likely to increase, rather than reduce, the need for comprehensive, independently verified financial product data. For advice firms increasingly competing with AI as a first port of call for clients, that is likely to mean leaning harder on the parts of the service an algorithm cannot replicate, rather than trying to out-automate it.
