FCA authorised · FRN 724952 0117 325 0027Quote & buy →
Aipedia: Claims management companies
AI makes it easy to fake documents, identities and claim stories, and CMCs that onboard automatically can pass them on to lenders and the Ombudsman. The firm is responsible for its checks, and PI is not designed for fraud losses.
Part of: AI in Claims management companies
In short
Generative AI can produce fake statements, agreements, medical letters and photos, and synthetic identities can pass basic checks. CMCs using automated onboarding risk sending fabricated claims to lenders and the Ombudsman, damaging genuine cases and relationships with partner firms. Under the Consumer Duty and the FCA's AI expectations, the firm is responsible for its controls. PI usually responds only to negligence claims, such as failing to spot fake evidence, and excludes the insured's own dishonesty. The firm's own fraud losses usually fall under crime or cyber cover.
Claims firms sit at the point where consumer stories become formal claims, which makes them a target. Generative AI makes it cheap to produce convincing fake bank statements, credit agreements, tenancy documents, medical letters and photos. Synthetic identities combine real and invented details to pass basic checks, and AI can write plausible accounts of an accident, a mis-sale or a housing defect in seconds. Bulk lead buying and automated onboarding mean some of this material can pass through to a lender, insurer or the Ombudsman before a person sees it.
No specific FCA rule on AI fraud for CMCs is set out in our sources, but the general position is clear. The Consumer Duty requires the firm to act in good faith, in PRIN 2A.2, and the FCA's October 2026 good-practice page says firms remain responsible for outcomes from AI they deploy, including tools such as automated onboarding. The FCA and SRA have also asked motor finance representatives to run onboarding checks for existing representation, as reported, which shows the direction of travel: regulators expect firms to know who their clients are and whether the claim is genuine.
Practical verification usually combines identity checks, liveness or document authenticity tools, checks that the claimant actually had the product or tenancy, and human review of anything that looks too neat.
A CMC presenting a complaint to a lender, insurer or the Financial Ombudsman Service is putting forward the consumer's case on their behalf. Submitting fabricated evidence, even unknowingly, damages the firm's credibility on every other case, can result in complaints being rejected, and may lead to regulatory questions about the firm's controls. The Ombudsman has already said it is seeing large, inaccurate AI-assisted submissions from professional representatives (see AI-written complaints to the Ombudsman). Where a firm suspects fraud, it should stop work, preserve evidence and take advice on its reporting obligations.
Illustrative scenario. A CMC buys leads for a new claim type and onboards them through an automated journey. A group of applicants upload AI-generated statements showing products they never held. The CMC's document tool extracts the figures and drafts complaints. Several lenders spot the pattern and reject every complaint from the firm in that batch, including genuine ones. A partner law firm then alleges the CMC breached its referral agreement by sending unverified claims.
Fraud can also run the other way: criminals impersonate the CMC with deepfaked voice calls or emails to divert a consumer's redress payment.
PI responds to negligence, not fraud. Many wordings exclude dishonest or fraudulent acts by the insured, although some protect the firm against the dishonesty of an employee or protect innocent partners (see the dishonesty exclusion). A claim that the CMC negligently failed to spot fake evidence, for example by a partner firm or a genuine claimant harmed by the firm's rejected batch, may be covered subject to the wording. The firm's own losses from fraud usually sit elsewhere: theft of money by deception may fall under crime insurance, and deepfake impersonation, account takeover and data incidents often fall under cyber cover, as our guide to cyber insurance and AI risks explains. Each policy has conditions, such as verification procedures, that should be checked.
Apex arranges PI for claims management companies, and a broker can explain how PI, cyber and crime cover might each respond to an AI-enabled fraud. See AI in claims management and Does claims firm PI cover AI errors?.
The fraud is the claimant's, but the firm may face complaints, regulatory questions and claims from partner firms if its checks were inadequate. The FCA says firms remain responsible for outcomes from AI they deploy, including automated onboarding.
PI usually covers third-party claims alleging the firm was negligent, for example in failing to spot fake evidence, subject to the wording. It does not usually cover the firm's own losses from fraud, which may fall under crime or cyber cover instead.
Losses from criminals impersonating the firm to divert payments or access systems are more often a crime or cyber insurance matter than PI. The answer depends on each policy's wording and conditions, such as verification procedures for payment changes.
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.
