By now, it is common to read that the COVID-19 pandemic accelerated digital transformation and innovation to enrich the customer journey and lower the cost of servicing.
While this is undoubtedly a win-win situation for customers and businesses in the era of social distancing, it has also created opportunities for anonymous fraudsters to exploit weaknesses in digital-first business models.
Digital transformation has made possible true 24/7/365 service over distance, as well as the capability to collect data and use it effectively to improve the customer experience.
Quick processing times have improved business efficiency markedly, but there is a reason 67% of claims and underwriting decision makers in the insurance sector believe there is now more fraud and a higher fraud risk than ever before.
That reason relates to customers’ lives moving increasingly online.
Via hacking and phishing, fraudsters have become adept at harvesting genuine customer details and using them in misrepresentation or synthetic identity creation to take advantage of quick processing times.
The rise of the ‘armchair fraudster’ is demonstrated in spikes in fabricated no-claims discounts, modification of new business quotes to reduce premiums or be accepted, phantom passengers, recycling of photos and statements of incident.
Fraudsters engage in exaggerated loss reporting, account claim takeovers and payment diversion that both distort underlying book data and skew premium calculation.
In short, as a typical example, fraudsters manage to extract low premiums and high claims from insurers, often hitting several underwriters in quick succession using the same details.
Fraudsters even pose as brokers, illegally attracting premiums, cancelling policies, diverting payments and falsifying documents.
Amit Kumar, Practice Leader, Fraud, AML & Security Intelligence, for SAS in Middle East & Africa, says trying to police and combat this type of fraud using only human capabilities is bound to fail.
“Identity fraud now represents more than 50% of all reported fraud and that figure is rising. Nine out of ten identity fraud incidents take place online and lead to mass applications for policies and claims. Trying to follow up once claims have been processed cannot stem losses.”
“The solution is data analytics as the first line of defence.”
SAS®Detection and Investigation for Insurance provides real-time data analytics that automates identity and claim verification throughout the claims life cycle.
“By quickly identifying claims activity as high-risk, the automated process can request further information and documentation for any flagged activity, using third-party data to provide additional defences.”
“More fraudulent activity can be detected, hidden relationships and linked identities can be uncovered and subtle patterns of behaviour can be detected at any stage of the claims process,” says Kumar.
The SAS insurance fraud analytics engine uses multiple techniques including embedded artificial intelligence (AI) and machine learning methods, to automatically score millions of claims records in real time or in batch.
Free-text, field-based or geospatial searches across all internal and external data can home in on particular network points for further scrutiny.
All details related to any particular investigation can be captured and displayed.
A consolidated view of fraud risk in a visualisation interface can reveal cross-product fraud and reveal social network diagrams, using sophisticated data mining capabilities to give insurers a better understanding of new fraud threats while preventing losses.
“Machine learning capability means the models continue to evolve and strengthen over time, staying on top of claims fraud trends,” says Kumar.
Loss prevention is not the only benefit of an automated fraud detection system.
“By quickly determining which claims require further scrutiny and which ones don’t, you can significantly reduce false positives – which means a better customer experience.”
“Automatic scoring lets you prioritise higher-value claims, entities and networks, while advanced case handling tools enable more efficient, effective investigations – and a higher return on investment per investigator,” adds Kumar.
“A biennial study by the Coalition Against Fraud Insurance and SAS has shown that by 2020 80% of more than 100 insurance companies polled use predictive modelling to detect fraud, up from 55% in 2018.”
“71% said they are considering investing in technology for the detection of claims fraud and 65% said they were looking to procure or expand their use of analytics to improve quality referrals.”
“It is clear that AI and data analytics are the only way insurers can counter rapidly changing fraud threats because criminals are using advanced technology at scale to steal personal information and commit fraud worth billions of dollars every year,” concludes Kumar.