Insurance companies increasingly use anti-fraud technology, which augments human intuition by claims staff. Mining of big data, for example, helps insurers to match patterns.
Something as simple as a vehicle identification number can highlight that the same vehicle has been involved in previous claims. This can save hours of human legwork.
Technology also catches out organised criminals thanks to geographical anomalies in the number of claims in a location.
Most Kiwis leave some sort of trail on social media these days. That can be used to check whether what they say in their claim rings true, says Phil Sylvester, from travel insurance company World Nomads.
"Sick in hospital? Then why is there a photo of you at the swim-up pool bar on the same date?" he says.
The company also records IP addresses automatically when policies are taken out, which can throw up fraudulent claims where the person is already overseas.
I once reported on a court case where a woman claimed to be too disabled to work.
She was followed by a private investigator who pictured her dashing across a road and later washing her car. She appeared in court in a wheelchair.
Ripping off the insurance company isn't a victimless crime because it costs us all in increased premiums.
These days a check of people's social media by investigators can be enough to catch such a claimant - which is a lot easier than paying a private investigator.
Southern Cross Health Society says as more claims are lodged electronically it becomes easier to use analytics and intelligence techniques to identify unusual or suspicious behaviour or data anomalies.
That includes fraud by healthcare providers, who may have asked patients covered by claims to sign blank forms.
Insurance lawyer Chris Boys of Assure Legal points out insurers have always looked for unusual activity or "red flags" to investigate.
"These can include things like attempting to sell the items on Trade Me before the alleged loss occurred," he says.
Fraudulent claimants were often caught through inconsistencies in their description of the loss.
"I've seen obviously forged invoices and arson claims where the insured's attempt to explain why a car and garage was burnt down was so fantastic that it couldn't have happened unless there was a vast, complicated and expensive criminal conspiracy behind it," says Boys.
Technology streamlines how easily these cases are identified and reduces the legwork needed.
For example, predictive modelling systems are now used to compare claims against known fraud indicators.
The system spits out numerical rankings according to how likely a claim is to be dodgy.
Human investigators can then focus on those the computer deems most likely to be fraudulent. Companies can also text mine - looking for inconsistencies in documents and notes, or language patterns in less than honest claims.
In New Zealand there is also the Insurance Claims Register, a database that most insurance companies check when claims are received.
The companies list details of claimants. Similar claims by the same person with one or more insurance companies will show up.
The register has a Hall of Shame. In one case when police raided the claimant's home they found many of the "stolen" items still there.
The red flag for that case was the policyholder claiming $10,000 on a $10,000 contents policy. It would be difficult for burglars to take every single item someone owned.
Ripping off the insurance company isn't a victimless crime because it costs us all in increased premiums.
In many cases it can really backfire.
If your claim is declined for fraud, including dishonest attempts at proving an honest claim, and it's listed on the Insurance Claims Register you will find it almost impossible to get any type of insurance in the future.
Of serious concern, that can mean you can't qualify for a mortgage because banks won't lend to people who can't get insurance on the house.
