UK insurers are facing a new fraud problem: evidence that looks ordinary at first glance but may have been edited, generated, or exaggerated with AI tools.
Aviva detected more than 18,400 suspect claims worth £233 million last year, a record total for the insurer. The figure includes brands Aviva acquired from Direct Line, making year-over-year comparisons less clean, but the pattern inside the claims is harder to ignore.
According to The Guardian, citing Aviva’s 2025 fraud data, scammers are using AI to fake car accident scenes, alter documents, and exaggerate damage.
Aviva’s UK general insurance business, excluding Direct Line brands, found that motor insurance fraud accounted for more than seven in 10 bogus claims.
Fraud is shifting from staged crashes to edited evidence
The fraud Aviva described is less about elaborate staged crashes and more about manipulated proof. The insurer said fraudsters are moving toward exaggerated claims for vehicle damage, repair costs, credit hire, and injury, with the value of detected motor fraud rising 39%.
That shift changes the job for insurers. Claims already move through photos, invoices, forms, and other digital submissions. When those materials can be cheaply edited or generated, fraud prevention becomes a verification problem, not just an investigation problem.
Aviva also reported more claims supported by AI-generated images and manipulated documents, particularly in motor insurance. The company said it is using AI tools and advanced analytics, overseen by people, to flag suspicious claims faster.
For businesses, the same risk extends beyond insurance. Forged invoices, fake documents, and altered screenshots can move through ordinary approval workflows if teams rely on surface-level checks.
Verification costs are becoming part of the claim
Aviva said fraud is also appearing in home and travel insurance, where some customers exaggerated the value of damage, repairs, or contents. Home insurance fraud across Aviva’s brands rose 15% in 2025.
The company also pointed to criminal consequences. Aviva said 37 years of custodial and suspended sentences were secured in 2025 for the most serious fraud offenses across Aviva and Direct Line brands. One case involved a deliberate collision and inflated claims worth £470,000.
The broader operational lesson is that digital evidence now needs more scrutiny. Insurers have the clearest numbers because claims create a paper trail, but the same verification burden can show up anywhere teams approve payments, refunds, benefits, expenses, or compliance filings from uploaded evidence.
That concern is already visible in AI-enabled scams, where attackers can use synthetic identities, voice cloning, and convincing impersonations to lower the cost of deception. It also explains why insurers are investing in AI use in claims processing while emphasizing human oversight, auditability, and compliance controls.
AI does not need to create an entirely new type of fraud to raise costs. It can make familiar fraud cheaper, faster, and easier to repeat.
For insurers, that means more investment in analytics, staff training, and claim-review systems. For other businesses, it is a warning that document-based workflows may need stronger checks before manipulated evidence becomes routine.
Also read: A study found ChatGPT Health missed some medical emergencies in test scenarios.