A Nottinghamshire man who filed four bogus claims for stolen Lego sets, MacBooks, and fishing gear — pocketing £14,000 from Axa Insurance — became one of the Association of British Insurers' headline fraud cases for 2025. Matthew Johnson was sentenced to 28 months in prison after City of London police found he had recycled the same receipts across different claims and doctored invoices to match policy addresses. The case is tidy and prosecutable, which is exactly why it became the poster child for a much larger and messier problem. The ABI's annual fraud report puts detected fraud at £1.34bn in 2025, a 14% jump on the previous year. The number of detected fraudulent claims actually fell 2.7% to 93,900, but the average value per case climbed to £14,300 — the second-highest on record, just below the 2022 peak of £14,600. Fewer scams, higher stakes. The industry's own language — "growing scale and sophistication" — acknowledges the problem is evolving faster than detection. Motor insurance remains the epicentre, accounting for 55% of all detected fraud. Some 51,900 fraudulent motor claims worth £625m were uncovered, with the average fraudulent motor claim hitting £12,000 — the second-highest since 2015. One case involved a man who staged collisions using women he met online in a classic "crash for cash" scheme. Aviva alone discovered £233m in bogus claims in 2025, reporting that scammers are now deploying AI to fabricate car accident scenes and supporting documents. Travel fraud showed the sharpest growth in case numbers. A single scam network used multiple fake identities, fabricated medical documents, and bogus websites to file false medical emergency claims worth £300,000. Exaggerated loss — deliberately inflating a legitimate claim — remains the most common fraud type at 26,900 identified cases, a reminder that the line between "honest customer" and "fraudster" is often a matter of degree. The extraction loop here is structural. Simon Klust of the City of London police fraud team says more than 250 bogus claims are detected daily but calls this "likely just the tip of the iceberg." The ABI and the Insurance Fraud Bureau — funded by insurers themselves — repeatedly frame fraud as a cost borne by honest policyholders through higher premiums. This is true, but it also obscures the incentive structure: insurers pass fraud costs through to customers, meaning the financial pain of undetected fraud lands on households, not insurer balance sheets. The Johnson case illustrates the detection model's limits. His fraud was unsophisticated — reused receipts, doctored invoices — and totalled £14,000 across four claims. He was caught because Axa's systems flagged documentation inconsistencies. The sophisticated, AI-enabled fraud the industry warns about operates at a different scale entirely, and the £1.34bn figure represents only what was caught. The actual fraud economy is almost certainly larger, and the gap between detected and actual fraud is the space where the real extraction happens. What the industry presents as a law-enforcement problem is also a market-design problem. Premiums absorb fraud losses as a cost of doing business. Insurers invest in fraud detection to manage — not eliminate — the problem. The Insurance Fraud Bureau exists to keep fraud at tolerable levels, not to drive it to zero. The honest policyholder sits at the bottom of this chain, paying higher premiums both for the fraud itself and for the detection apparatus built to contain it.
