Why AI financial advisers have a leg-up on their old-world rivals

Search engines weren’t designed to be diagnostics businesses, but millions of people consult “Dr Google” before seeing a real physician. Artificial intelligence is having a similar effect on personal finances. General purpose chatbots are increasingly used for financial advice. Call it ChatIFA.

Already, almost a fifth of UK consumers use AI to help with personal finances, according to a report this week by the Financial Conduct Authority. They’re not just summarising information: 61 per cent of AI users said they asked for suggestions and almost a quarter upload personal data such as bank statements for better answers.

The risk to financial firms is obvious: if AI gives sophisticated, personalised recommendations for free, why pay an expensive adviser? Customers may still need a broker to act on the robot’s recommendations, but that’s not where the profit is in financial services. Such fears have recently upset the shares ofmass-market wealth managers like Charles Schwab and Raymond James in the US and St James’s Place in the UK.

This isn’t a simple case of fusty incumbents disrupted by novel technology. Established groups can easily outdo the offerings of chatbots: they have big enough tech budgets to build digital advisory interfaces and masses of data to inform the suggestions. They could even be more convenient: no need to upload statements to a bank that already knows your salary, spending and saving habits.

The catch, of course, is that regulators won’t let them. Orthodox financial firms can’t hand out personalised advice willy-nilly; there are strict rules to protect consumers, and punishments for getting it wrong.

One option is to label AI-powered wisdom as being different from the old-fashioned kind, and hope that investors know or care about the difference. Banks such as Lloyds and Barclays are working on tools to provide “targeted support” — a new midpoint between specific “advice” and generic “guidance”. But that still involves carefully calibrated information. Google’s Gemini, in contrast, will confidently respond to scant inputs of personal information for a UK saver by saying “your absolute priority should be a Lifetime ISA”.

It is, though, too much to expect mild disclaimers to allay the risk of bad counsel. Large language models are designed to sound convincing and humans often over-trust AI outputs. Only 40 per cent of respondents in the FCA survey realised there was no way to complain if something goes wrong after consulting AI about their finances.

The FCA at least noted the risk of an “uneven playing field” between regulated firms and tech platforms. Don’t expect any swift action: one of the review’s key recommendations was another review. Chances are the watchdogs will intervene eventually, though. That could involve prominent, tobacco-style warnings when British users ask chatbots what to do with their nest egg — or the requirement to direct curious users to real, licensed advisers.

There’s another kind of uneven playing field to watch for, too. If the UK ends up tougher on financial chatbots than other countries, it could make the country’s wealth managers look like a more attractive investment than their peers in less restrictive regions. Executives often complain that overzealous regulation holds them back; in this case, cautious rule-setters might actually give them a boost.

[email protected]