The growing threat of AI-Enabled Fraud in large loss claims

The use of Artificial Intelligence (AI) in detecting fraud is now integral to insurer and lawyer processes. A range of tools are used in personal injury claims, from triaging identifying potential fraud to using AI to identify inconsistencies in evidence and social media analysis. Metadata analysis is standard for those working in this field.

Becoming more prevalent is fraudsters using AI in fake supporting documents or to alter images or invoices. It does not take a specialist to alter photographs to create fake accident images or fake images of accident locus, and even injury images. The fraudulent presentation of claims is frequently seen in the large loss space. Although the recent high profile Shaw v Wilde defendant win where the claimant’s £6.6 million claim was struck out because he was fundamentally dishonest did not involve AI created fraud, it seems that claimants are increasingly emboldened to present larger claims, or aspects of claims, which are fraudulent.

According to a recent report, one insurer said they have seen a 300% increase in incidents where real life images, videos and documents have been manipulated. Similar trends are reported by other insurers and in that same report, 83% of claims handlers surveyed suspected that AI was used to alter 5% of claims.

Periods of fraudulent activity always spike when there is cost pressure on families, and the current financial backdrop can mean a greater temptation to present a claim that has a fraudulent element or is entirely fraudulent. With detected fraud in the UK said to cost over £1 billion each year, the estimate for undetected fraud is double that.

Although it is still the case that fraudsters tend to focus on smaller value claims, there is plenty of scope for fraud, including AI-created, in the large loss sphere as well. Our fraud specialists work closely with our large loss lawyers on these cases. There is still no substitute for experienced instinct in detecting fraud as well as deploying the right defence strategy.

We know that the audio and visual manipulation, or deep fakes, is a major growth area for insurance fraud but the skill still lies in identifying those cases that demand further investigation.

Insurers’ and lawyers’ fraud detection systems continue to evolve at pace to keep ahead of the fraudulent claimant and, of course, AI is integral to that response.

A cautionary tale: AI is not always the answer – not appreciated by the New York lawyer who in 2023 submitted court filings referencing cases proposed by ChatGPT that did not exist. Even though he again used ChatGPT to double check and asked “is this a real case?”, to which ChatGPT replied “yes”, unsurprisingly, this was not a sufficient defence in his own misconduct hearing.