In a recent study by FRISS, the worlds most implemented AI powered fraud, risk and compliance solution provider for P&C insurance companies worldwide, the challenges and opportunities facing insurers in their efforts to combat fraud throughout the entire policy lifecycle are highlighted. The study gathered input from over 400 insurance professionals worldwide and provides insight into topics such as fraud schemes, data challenges, process automation and more.
Survey respondents have differing views on the challenges and benefits of fraud detection software solutions. The common theme however is the data challenge; from underwriting to claims to special investigations. The difficulty is harnessing timely data to respond quickly when fraud is detected. FRISS’ past biennial surveys indicate that insurance professionals have struggled with inadequate data – either poor-quality internal data or limited access to external data sources.
Key findings of the fraud study include:
The Pandemic accelerated digitalization
COVID-19 will have a lasting impact on insurance, in large part because the pandemic has accelerated digital processes. Along the lines of studies by EY and Aite-Novarica, insurers must continue to address their technology debt by digitizing core processes, migrating to the cloud and embracing flexible sourcing models. Insurers are using a multilayered approach to minimize their risk of fraud, and improving protection at the front door at underwriting is definitely an area of focus. While these trends are continuing, insurers also are better positioned to take advantage of digital tools to combat fraud end-to-end.
Fraudsters remain creative
Exactly how much fraud impacts the industry is always hard to pinpoint, however according to the Coalition Against Insurance Fraud, in the US alone, fraud steals at least $80 billion every year from American consumers. Creativity and persistence in claims fraud is therefore a serious threat for insurers. Fraudsters continue to perpetrate schemes against insurers and inevitably work to exploit emerging system gaps, continuing to drive up the cost of insurance for honest consumers. For 41% of survey respondents, keeping up with modern fraudster modus operandi was their greatest challenge in effective responding to fraud. The top fraud schemes that saw an increase in popularity the past year are claiming false injuries, nondisclosure of relevant information and staged accidents.
The industry generally agrees fraud accounts for about 10% of all claims cost. However, one change since the prior survey is an increase in the percentage of claims suspected as fraudulent. In 2021 the suspicion of claims containing a potential element of misrepresentation or fraud rose to 20% – a rise predicted by FRISS in the previous report.
Data crucial in fraud fighting
Having the right data in the right place, and in real time, is essential to improving fraud detection. With many insurers utilizing digital processes for almost all of their operations, the ability to see real-time data identifying potential fraud is hugely beneficial across the policy lifecycle – from first-party policy requests, to underwriting, and of course as claims are reported. The difficulty is harnessing timely data to respond quickly when fraud is detected. FRISS’s past biennial surveys indicate insurance professionals have struggled with inadequate data – either poor-quality internal data or limited access to external data sources. This year, amongst the top challenges in fighting fraud again where data protection and privacy, internal data quality and inadequate access to external data.
Optimization is upon us
The future of fraud detection lies in the use of advanced technologies to support real-time, large-volume, and highly precise modelling for claims and underwriting fraud. Fortunately, respondents do see significant benefits in fraud detection software. These include:
- Improve loss ratio, cited by 59%
- Stay ahead of developing fraud schemes, cited by 53%
- Increase investigator efficiency, cited by 52%
A hybrid approach of human expertise and predictive models will be essential in preventing losses. This will reduce the costs of underwriting and claims handling by removing unnecessary and error prone steps, and enables discovery of suspicious behavioural trends in data. This not only can augment the results of the existing data, it would give insurers an advantage when identifying the ever-changing schemes of fraudsters. Because FRISS believes that when insurance is more transparent and everyone can pay fair premiums that aren’t inflated by the real costs of fraud, businesses and individuals can thrive and achieve their dreams.