Zurich UK says connected AI analytics found £17.9m of fraud within a record £109.3m year

Zurich UK credits connected AI and analytics with uncovering 860 fraudulent claims, worth £17.9m, over the past 12 months, or roughly a fifth of every fraud case it caught, according to figures published by Insurance Business on 5 November. The system runs on Quantexa technology and checks each claim against Companies House and Equifax records, industry intelligence and the Insurance Fraud Register. Zurich describes itself as the first insurer to join up every claim across several data sources for a single claimant.

Joining the records up revealed people who had used other names to conceal past claims, claims that looked unrelated but shared one accident location, and groups of rogue solicitors and accident management firms working together. Zurich says the checks also help honest customers: when a claim scores as low fraud risk, its handlers can pay it sooner.

Taking all detection methods together, Zurich UK says it found £109.3m of fraudulent claims, its highest value and number on record. The 4,359 cases compare with 3,562 in 2022, a rise of 22%. By value, casualty claims made up £61.6m, motor £28.1m and property £19.6m.

Scott Clayton, who leads claims fraud work at Zurich UK and chairs the Insurance Fraud Bureau, says the criminal networks keep changing tactics, from staged accidents and ghost broking to invented claims and evidence produced with AI. He called the mix of analytics, specialist investigators and human judgement a strong defence.

The insurer's figures sit within a rising market total: the Association of British Insurers says UK insurers detected £1.34bn of fraudulent claims in 2025, 14% more by value, although the number of cases fell slightly.