FCA authorised · FRN 724952 0117 325 0027Quote & buy →
Aipedia: Wealth managers and discretionary fund managers
AI can help wealth managers spot signs of vulnerability across large client books, but flags only reduce risk when someone acts on them. The FCA found most firms identify vulnerable clients, yet only around 36% of those adjusted their service.
Part of: AI in Wealth managers and discretionary fund managers
In short
The FCA's 2026 survey found 83% of discretionary firms identified at least one vulnerable client, but only around 36% of those adjusted services, with inconsistent practice. AI tools can scan calls, emails and transactions for signs of vulnerability or financial abuse, but they process sensitive data and can miss how vulnerability shows up in older clients. The Consumer Duty and the FCA's October 2026 good-practice page expect outcome testing and non-AI options. Missed signs usually lead to ombudsman complaints, which PI commonly covers subject to the wording.
Discretionary clients are often older, wealthy and long-standing, which brings particular vulnerability risks: declining capacity, bereavement, serious illness, and pressure from family members or attorneys. The FCA's Wealth management survey report 2026 found that 83% of firms identified at least one vulnerable client in 2024/25, up from 68% in the first survey. But only around 36% of those firms had adjusted services for vulnerable clients, and the FCA describes practice as inconsistent.
AI tools can scan call transcripts, emails and meeting notes for signs such as confusion, repeated questions, references to illness or bereavement, or a third party who starts speaking for the client. Others flag unusual transaction patterns: large withdrawals, new payees, or a sudden change in investment instructions. Flags go to the portfolio manager or a vulnerability champion for follow-up.
These tools can extend coverage across large books, but they also process special category data such as health information, which requires a lawful basis and care under data protection law; see the ICO guidance on AI.
The main risks are missed flags and flags that are raised but not acted on. A tool trained mostly on younger or digital clients may not recognise how an older client's vulnerability shows up. An attorney or relative may make plausible requests that the tool treats as normal because they come from an authorised person.
Illustrative scenario: a widowed client's son, acting under a power of attorney, asks the firm over several months to move the portfolio to a higher-risk strategy and to make regular payments to an account in his name. The AI monitoring tool flags the payments, but the alerts are closed as authorised by the attorney. The client's other children later complain that the firm failed to notice signs of financial abuse and allowed an unsuitable change of strategy.
Complaints about missed vulnerability usually reach the Financial Ombudsman Service, which decides on what is fair and reasonable and will consider whether the firm followed FCA expectations. If the firm is required to pay compensation for negligent service or an unsuitable portfolio, PI usually responds, subject to the wording and excess. Claims alleging the firm should have detected another person's dishonesty can raise questions about fraud exclusions, so check how they are drafted. The ombudsman says it receives very few complaints about AI itself, so the complaint will usually be framed around the outcome, not the tool.
Evidence matters. A record showing that a flag was raised, who reviewed it, what they checked and why they decided as they did is far more persuasive than an alert log showing flags closed without comment. Firms that cannot produce that record risk the ombudsman concluding that the tool was in place but the firm did not use it.
Apex arranges PI for wealth managers and discretionary fund managers. A broker can check how your policy treats ombudsman awards and claims involving third-party abuse of a client's account. See AI in wealth management.
No. The FCA's survey shows identification is common but service adjustment is not. A flag only helps if someone acts on it, reassesses suitability where needed and records what was done.
An attorney's instructions are generally valid, but the firm still needs to be alert to signs that they are not in the client's interest. Treat unusual requests from attorneys as a reason to look more closely, not as automatic clearance.
PI usually responds to ombudsman awards arising from negligent professional service, subject to the policy terms and excess. Check exclusions relating to fraud by third parties and whether goodwill payments are covered.
Tell us about your firm and the work you do. A broker will review how your current PI treats AI-assisted work, or arrange new cover. No fee for a quote. Or call 0117 325 0027.
Start a PI quote Request a callbackApex Insurance Brokers Limited is authorised and regulated by the Financial Conduct Authority. Registered in England and Wales, company number 07014570. This page is general information about how AI is used in the profession and how insurance may respond. It is not legal, regulatory or insurance advice on your circumstances, and it does not guarantee that cover will be available or on what terms.
