UK brokers unprepared for AI-driven insurance fraud crisis
Insurers are blocking hundreds of millions in fraudulent claims, yet most brokers still underestimate the threat. Allianz stopped 32,400 fraud cases in 2025, preventing £174 million in payouts-up from £157 million in 2024. Criminals used generative AI tools to doctor vehicle and property images pulled from social media, then submitted fake repair invoices to support third-party claims.
In one case, a business owner's van copied from his company's social media page was given an AI-generated crack on the bumper and used to claim £1,000 in repairs that never happened.
The cost falls on honest policyholders
Matt Crabtree, head of financial crime at Allianz, said undetected fraud would ultimately raise premiums for genuine customers. "If these attempts keep increasing and some go undetected, it could affect premiums for genuine customers," he said.
The Association of British Insurers recorded £1.16 billion in detected fraudulent claims across 98,400 cases in 2024. The ABI estimated fraud adds around £50 a year to the average policyholder's premium. Admiral detected nearly £87 million in fraudulent claims in 2024, up 71% from £51 million the prior year.
Fraudsters now generate fake policies
The threat extends beyond manipulated images. Criminals are now generating wholly fictitious policy documents to undercut legitimate brokers on price. This leaves consumers uninsured while honest intermediaries compete against fraudulently low premiums that carry none of the costs of underwriting or regulation.
The Insurance Fraud Bureau tracked a 52% rise in ghost broking cases between 2022 and 2024. FCA research from May 2026 found that 49% of young drivers had purchased insurance through social media or messaging platforms. Four in 10 said they would not feel confident identifying a fake policy.
Brokers lag behind the threat
Yet most brokers do not view AI as a pressing concern. GlobalData's 2025 UK Commercial Insurance Broker Survey found that only 5.2% of brokers see AI as the biggest threat to their business. This sits in sharp contrast to the fraud data and growing SME interest in AI-driven insurance services.
A corporate criminal offence under the Economic Crime and Corporate Transparency Act 2023 came into force on September 1, 2025. The law holds organisations liable for failing to prevent fraud by associated persons. While the direct scope targets only the largest companies, smaller firms throughout supply chains are not immune.
What brokers and customers should do
Matt Crabtree urged policyholders and brokers to remove identifying details from any images posted publicly. "If you are posting pictures of your cars, watches or technology online, remove any personal identifiable details like registration or serial numbers," he said.
Nicola Smith of the Insurance Fraud Bureau said the pace of change had outstripped most people's expectations. "Criminals are more sophisticated than ever, and with the rise in use of AI, they exploit information in ways most of us wouldn't expect," she said.
Brokers managing AI for insurance operations need to understand both the fraud methods and the regulatory obligations now in place.
Your membership also unlocks:
