AI in UK Wealth & Investment Advice

Why this briefing

CP26/10, published on 25 March 2026, proposes to remove the annual suitability requirement in favour of periodic reviews based on clients' needs, and to replace the requirement to consider "necessary" information with an expectation that advisers consider "sufficient" information. The FCA is explicit: firms can use "technology to drive innovation in both how they assess suitability on an ongoing basis and deliver the suitability review".

The review is what advised clients pay for. About 80 per cent of adviser-charge revenue comes from ongoing services, from around 3.7 million clients. The FCA's 2025 review found reviews delivered in around 83 per cent of cases at 22 large firms — and two of the largest advice businesses report, in their own results, provisions still open for reviews they could not evidence.

Assessing suitability on an ongoing basis means something has to notice when a client's life has changed enough to need an adviser. In a firm with thousands of clients, that is a trigger — increasingly a model. And the capability is not evenly spread: 95 per cent of large advice firms are using or considering AI in the advice process, against 48 per cent of small firms.

The human moment

The suitability review — and in particular the review that, under the proposed rules, happens only when something prompts it. Has this person's life changed enough to need an adviser? Under the annual rule the calendar answered that question. Under the proposed rule, the firm answers it for every client, every year.

What's inside

Sixteen pages, written for an executive who knows the sector well and AI not at all:

• A note on the numbers — what is included, what was excluded, and why
• A trillion pounds under advice, and a workforce that has not grown
• The pressure point — advice reaches fewer than one adult in ten
• The review loses its date, and the fact-find loses a word
• Four risks specific to advice, none of them about the technology working
• What the record shows — reviews delivered, and what unevidenced ones cost
• The threat side — the unregulated adviser already in the client's pocket
• The UK legal position, and the change from "necessary" to "sufficient"
• Beyond the UK — the European supervisor's AI statement for advice
• Seven things that can be done now, none needing new law
• Three levels defined by where accountability sits, and three horizons
• An eight-workstream way in

The discipline behind it

Every figure carries a source: the FCA's consultation and market surveys, the Financial Ombudsman, the European Securities and Markets Authority, UK Finance's own data, or a firm's own published results. Where a firm's remediation provision is used, it is the firm's own figure and no conclusion beyond what the firm has published is drawn.

Robo-advice market forecasts, adviser time-saved claims from note-taking tool suppliers, and a widely quoted failure rate for AI advice produced by a competing firm are all excluded on the record. So is a government AI-use percentage quoted in the FCA's own consultation, because its underlying survey could not be read.

Who it's for

An advice firm or wealth manager executive who has just picked up an AI programme, a compliance or advice-quality lead, or a board member being asked to approve AI in the fact-find or the review. It assumes advice is well understood and AI is not.