UK insurers detected £1.16bn in fraudulent general insurance claims in 2024, across 98,400 cases - a 12% jump year on year. The spread of artificial intelligence is adding a fresh layer of risk: synthetic fraud, where fabricated digital evidence and fake identities can be created with unprecedented speed, putting pressure on claims teams already processing thousands of submissions each week.
AI changes the fraud playbook
Synthetic fraud - using artificially created documents, media or identities to invent or inflate a claim - is not a new tactic. But Matt Gilham, director at WhiteElk Fraud Performance Consulting, said the technology has shifted the threat. "What has changed is the speed and ease with which fraudsters can find, create or amend digital evidence. However, the true underlying frequency of fraudulent evidence, particularly documents, is still not consistently quantified," he said.
Jon Bethell, head of private clients at Verlingue, called the rise of AI in claims fraud "the most heavily underestimated threat the market is facing." The tools are simple to misuse. "People are using AI to create fraudulent claims. An example of this would be, you take a picture of your living room and you ask AI to show that there's been an escape of water and the roof's cracked," he explained.
Fake receipts, real payouts
Not every synthetic claim requires a dramatic fabricated scene. Bethell pointed out that claims inflation can be as basic as generating a fake receipt for work that never happened, using widely available AI tools. These tools, driven by Generative AI and LLM technology, have lowered the barrier to entry for would-be fraudsters. A busy claims department handling thousands of claims a week can easily miss a single forged document mixed into a legitimate submission.
"If you fire a fake AI receipt into every single one of those claims, [that fake evidence] will get through because they're really, really busy," Bethell said. The economics of verification don't help. While high-value cases will always see a loss adjuster, sending an inspector to every £2,000 claim isn't viable.
Insurers face a detection challenge
Insurers are responding with better image verification tools and are exploring ways to validate documents and video. But Gilham warned that these solutions have their own weaknesses. Insurer processes can strip key metadata from submitted media, weakening the signal that detection models rely on. False positives and false negatives still need human oversight, adding to the workload of teams already managing fraud indicators from multiple sources.
A newer risk is also emerging: claimants challenging insurer-gathered digital evidence as AI-created or altered. "We are seeing the emerging risk of claimants challenging insurer-obtained digital evidence as AI-created or altered, increasing the need for a stronger chain of evidence for digital media," Gilham said.
Fighting synthetic fraud with AI
The same technology can be turned back on fraudsters. Insurers are starting to deploy AI for Insurance fraud detection, training models to spot doctored images, duplicate submissions, and metadata inconsistencies. Text-based claims can also be screened for signs of AI generation rather than human authorship. These defensive tools aren't a cure-all, but they offer a way to match the speed and scale of synthetic fraud attempts.
Why this matters for claims and fraud professionals
For claims teams and fraud investigators, the message is clear: digital evidence can no longer be taken at face value. A photograph, a receipt, or a repair quote might be entirely fabricated by AI in minutes. The operational implication is a need to cross-reference submissions more thoroughly and to invest in AI-assisted detection that keeps pace with the tools fraudsters already have. Underwriters and loss adjusters who ignore this shift risk inflated loss ratios that erode portfolio performance quietly, claim by claim.
Your membership also unlocks:
