Professional criminals and amateur ‘chancers’ alike are increasingly likely to fall foul of new digital detection techniques.

In 2017 the Association of British Insurers reported that the value of detected fraud amounted to £1.3 billion. However, despite the increasingly sophisticated activities of insurers to take the fight to fraudsters, it is clear that this figure is dwarfed by the amount of undetected fraud – and that paradoxically, it is not the insurers themselves who directly suffer the consequences of this criminal activity.

Instead, it is the ordinary policyholder who pays more in annual premiums than would otherwise be the case. By some estimates this is thought to be some £50 per year for every motor insurance policy issued. The honest customer continues to subsidise the criminal fraternity.

That is not to say that insurers are washing their hands of the problem. Quite the opposite. Prevention and detection systems continue to grow in sophistication. From the initial policy application to the claims notification and settlement process, insurers are investing heavily in new systems of information collection and data analysis.

Many of these systems revolve around two core principles.

Firstly, wherever possible, data should be collected and analysed in real time. During the relatively short period of notifying a claim (perhaps just a few minutes, especially if online facilities are being used), the insurer needs to know if a claim is potentially fraudulent. Data from hundreds of sources needs to be automatically collected, examined and reported so that the claims handler or automated system can react accordingly – all in the blink of an eye.

Secondly, the data being collected and analysed will be a mass of diverse information that needs to be rapidly examined for linkages in claim circumstances, claimant details or other red-flag factors.

For example, a motor damage incident reported as having occurred in a car park, late at night, and in a location previously viewed as being suspicious would immediately raise a red flag for the insurer to examine further. Add to this the potential for identifying if the individual involved is connected to other confirmed or suspected fraudsters, perhaps by something as apparently minor as a common mobile phone number across two different claims submitted months apart, and the web of connectivity and suspicion spreads far and wide and at a very rapid pace.

Sophisticated software can automatically trawl social media, publicly available data, specialist fraud records, claims history, court records and a vast panoply of additional information that is growing by the day, and which makes fraud detection ever more likely to be successful.

However, the professional fraudsters know all this, and it is a constant running battle for insurers and authorities to stay ahead of the game.

Nor is it just the hardened criminals that contribute to the cost and consequences of fraud. Surveys have suggested that up to 25 per cent of household insurance claimants either have, or would consider, artificially exaggerating the value of their otherwise legitimate claim. Insurers view the personal injury market as ripe for fraudulent activity, with over 700,000 claimants anually. Sadly, there are still those in the legal and medical professions who facilitate attempts to commit fraud.

These factors and others mean that while the new digital world of fraud prevention and detection is at an all-time high in terms of ability to thwart the aim of professional and amateur alike, it is a battle that is by no means over, and where the real victim is the ordinary policyholder who is paying too much for their insurance.