FCA authorised · FRN 724952 0117 325 0027Quote & buy →
Aipedia: Wealth managers and discretionary fund managers
Wealth managers can use AI to draft quarterly reports and commentary, provided every figure and holding is checked against source records. Clients rely on these documents to decide whether to stay invested, so misstatements can lead to complaints and claims.
Part of: AI in Wealth managers and discretionary fund managers
In short
Generative AI can produce quarterly narratives and market commentary quickly, but it can also invent explanations, list sold holdings or report the wrong period's performance. The Consumer Duty's consumer understanding outcome and good faith obligation apply to every report. A client who relies on a misstated figure and loses money may bring a misstatement claim, which PI commonly covers subject to the wording; reissue costs usually are not. The safest workflow takes figures from the firm's systems and uses AI only for narrative.
Quarterly valuation packs, market commentary and performance narratives are obvious candidates for generative AI. A typical workflow feeds the tool with holdings, transactions, performance data and the investment committee's notes, and asks for a client-ready summary in the firm's house style. Some firms personalise commentary per client; others generate one market view and a short portfolio-specific paragraph.
The output is fluent, which makes errors hard to spot. AI can invent a reason for a holding's performance, describe a sale that did not happen, or mix figures between periods or share classes.
The commonest errors are performance figures for the wrong period, holdings listed that were sold, commentary explaining a market move the portfolio did not participate in, and charges summarised wrongly. Generic commentary can also contradict the portfolio, for example praising a sector the client's mandate excludes. These are forms of AI hallucination.
Personalised commentary adds a confidentiality risk. If the tool is given several clients' data in one session, it can carry details from one client into another's report. Sending a client someone else's holdings or circumstances is a data breach as well as a service failure, and the ICO's guidance on AI and data protection covers assessing risks like this before deployment. Batch generation also means one faulty template or data mapping can affect every report in a quarterly run, so a single error can become a large remediation exercise.
Illustrative scenario: a firm's AI tool generates quarterly narratives for a batch of clients. A data mapping error causes it to report the performance of a higher-risk model to clients in a lower-risk model. A client relies on the figures, decides not to move their portfolio and adds new money. When the error comes to light months later, the client complains that they were misled and would have acted differently.
A misleading statement in a client report can found a claim for negligent misstatement or breach of contract, and PI commonly responds where a client can show reliance and loss, subject to the wording and excess. Many complaints will not show a loss: the client is angry but the portfolio itself was suitable. Costs of correcting and reissuing reports, and goodwill gestures, are usually for the firm unless the policy has a specific extension.
Where a batch error is discovered, consider whether it is a circumstance that should be notified; see circumstance notification explained. Complaints about clarity under the Consumer Duty also interact with PI as discussed in Consumer Duty and PI.
Apex arranges PI for wealth managers and discretionary fund managers. A broker can review how your policy treats misstatement claims, reissue costs and notification of batch reporting errors. See AI in wealth management.
No specific rule requires it, but the FCA's October 2026 good-practice page encourages telling consumers how AI is used and its limits. Any description should be accurate and not overstate the role AI plays in investment decisions.
Only if it leads to a claim or complaint alleging loss. A client who relied on a misstated figure and lost money may claim, and PI commonly responds subject to the wording. Many errors are corrected without any loss.
Treat it as a potential personal data breach and assess whether it must be reported to the ICO. Any resulting claim may fall under cyber cover, PI or both, depending on the wordings, so notify the relevant insurers promptly and keep a record of what was sent.
It is safer for figures to come from the firm's performance and custody systems, with AI used only to write narrative around verified numbers. That separates calculation from drafting and makes errors easier to trace.
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.
