Detecting Smartphone Insurance Fraud

Sammanfattning: The vast majority of Swedes are in possession of a smartphone, typically covered via one’s household insurance policy and smartphones are included in upwards of 50 % of household insurance claims (Hedvig, 2019). Some estimates say that upwards of 40 % of all smartphone claims are fraudulent. Payments to fraudsters falls on honest policy holders pay for with higher premiums. There are several vulnerabilities in the current measures of Insurance Fraud Detection (IFD) on smartphones regarding the indicators that invoke suspicion and the economic incentives for investigation. The purpose of this report was to identify viable Smartphone Insurance Fraud Indicators (SIFIs) to be used for IFD, along with suggestions for further improving the fraud detection capability in areas of data acquisition and investments in analytical tools. A triangulation methodology was employed, which extended to interviews with 12 practitioners of IFD, attendance at 2 international insurance conferences and a review of 20 published academic papers and articles on insurance fraud. The primary result of this report was a compilation of 51 distinct SIFIs. By contrasting and comparing the findings from the three method areas, several examples of divergences between theory and practice were identified. The results can be used to expand and revise existing sets of indicators, as well as prioritising investments in new analytical tools. Although the research is focused on smartphone claims, the results also have the potential to be apply to other common belongings such as laptops, and tablets.