FCA authorised · FRN 724952 0117 325 0027Quote & buy →
Aipedia: Insurance brokers
An AI-drafted market submission must still give the insurer a fair presentation of the risk under section 3 of the Insurance Act 2015. If the draft omits a material fact or invents one, the client may lose cover and the broker may face the claim.
Part of: AI in insurance brokers
In short
Brokers use generative AI to turn client information into risk presentations, proposal forms and market submissions. The Insurance Act 2015 requires the insured to disclose every material circumstance it knows or ought to know, clearly and accessibly, with material factual representations substantially correct. A model that leaves out a loss, smooths over a known issue or invents a plausible detail can give the insurer a remedy that reduces or defeats a claim. Because the broker prepared the presentation, the client is likely to look to the broker, and the broker's E&O cover, for the shortfall.
Preparing a good submission takes time: collecting the client's data, writing a narrative of the business, summarising claims history and answering insurer questions. Brokers now use AI to draft the narrative from the client's website and documents, populate proposal forms from previous years, summarise loss runs and adapt one submission for several markets. Some tools also answer insurer queries by searching the broker's file.
Generative models write fluent, confident text. That is useful for presentation, but it can hide the fact that a statement was inferred, out of date or simply made up.
Under section 3 of the Insurance Act 2015, before a non-consumer contract is entered into the insured must make a fair presentation of the risk. That means disclosing every material circumstance the insured knows or ought to know (or enough to put a prudent insurer on notice to ask), in a manner reasonably clear and accessible to a prudent insurer, with every material representation of fact substantially correct and every representation of expectation or belief made in good faith. The Insurance Act 2015 overview explains the remedies when the duty is breached.
On the regulatory side, COCON 2.1.2R requires due skill, care and diligence, and Principle 7 in the FCA Principles, which the FCA AI Update lists, covers communications. Under SYSC 9.1.1R the firm must keep orderly records, which should include what was sent to each market.
Illustrative: a broker uses an AI tool to refresh a client's liability submission from last year's file. The client mentioned on a call that it had started roof work at height, but the model drafts from the old file and describes the business as "ground-level installation only". The insurer quotes on that basis. After an injury claim, the insurer finds the statement was wrong and argues it would have charged a higher premium or declined. The client faces a reduced settlement and blames the broker.
Other risks include an AI summary of the loss history that drops a small claim, a "dressing" of facts that makes them look better than they are, and a proposal answer copied from another client's file.
Where a client's claim is reduced or avoided because of a defect in a presentation prepared by its broker, the client will often bring a negligence claim against the broker. A broker E&O policy meeting MIPRU 3.2.4R would usually be expected to respond to that kind of claim, subject to its terms, any excess and any AI-specific conditions. The broker's position is stronger if it can show the client approved the final presentation and was asked to check it.
The same discipline applies to the broker's own PI proposal; see the fair presentation duty.
See also AI data enrichment and prefill, AI market selection and placement and the insurance broking AI hub.
Apex arranges PI for insurance brokers. A broker can check how your E&O policy responds to presentation errors and whether insurers expect any controls over AI-drafted submissions.
Yes. The Insurance Act 2015 duty sits with the insured, and the way the presentation was drafted makes no difference. If an AI-written submission omits a material circumstance or contains an incorrect material fact, the insurer may have a remedy.
The duty to the insurer is the client's. But where the broker drafted the presentation and the error came from the broker's process, the client may sue the broker in negligence. Getting the client to review and confirm the final version helps to share responsibility.
It can, if used to prompt rather than to write. For example, a tool that checks a draft against a list of material circumstances for the trade, or flags inconsistencies with the client's documents, can catch gaps a person might miss.
Keeping the final version sent to each insurer is essential under SYSC 9.1.1R record-keeping. Keeping the source documents and the client's sign-off is also valuable, because it shows where each statement came from if the presentation is later challenged.
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.
