Aviva detects record £233m in fraud as AI scams rise

TL;DR:

• Aviva detected more than 18,400 suspect claims worth £233m in 2025 — a record, boosted by its Direct Line acquisition.
• Fraudsters increasingly used AI to fabricate accident scenes, documents and damage imagery, particularly in motor insurance.
• Aviva is countering with AI analytics overseen by humans to flag suspicious claims faster.

The arms race between insurers and fraudsters has a new front: generative AI. Aviva detected bogus claims worth more than £233m last year, a record for the insurer, as scammers turned to AI tools to fake car accidents, manipulate documents and exaggerate damage. The firm identified over 18,400 suspect claims across its brands, the first full year including the Direct Line brands it acquired last summer.

AI on both sides of the claim

Motor insurance accounted for most detected fraud — more than seven in ten cases in Aviva’s UK general insurance business — with the value of motor fraud up 39%. The insurer said fraudsters were shifting away from staged collisions toward exaggerated claims for damage, repairs and injury, increasingly “supported by AI-generated images and manipulated documents”. Pete Ward, head of claims counter-fraud, warned fraud “isn’t a victimless crime — it drives up the cost of insurance for everyone”, and said scams were becoming “more sophisticated”.

The response is symmetrical: Aviva is using AI tools and advanced analytics, kept under human oversight, to detect suspicious claims faster. The stakes are concrete — in one case, fraudsters staged a collision to claim £470,000, with court video showing none of the witnesses had been present, leading to convictions. Aviva secured 37 years of custodial and suspended sentences in 2025, and also reported a 15% rise in opportunistic home-insurance fraud.

Looking forward

Aviva offers an early, quantified look at AI-enabled fraud hitting a major UK sector — a pattern likely to spread to banking and the public sector, echoing warnings that AI is arming low-skill bad actors. The contest will hinge on detection keeping pace with fabrication, and on the human oversight insurers stress to avoid wrongly rejecting genuine claims. For UK consumers, the immediate concern is cost: as synthetic evidence proliferates, the price of policing it risks feeding into premiums.