According to FBI reports, insurance fraud including bogus claims and shady brokers cost insurance companies and their clients over $40 billion per year. This figure excludes medical insurance fraud, which industry groups say add tens of billions to the cost. The feds added that the increased level of fraud affects all people, mainly by including $400-$70 per year to the premiums all of us pay to healthcare, cars, and homes.
Since the number of investigators is not sufficient enough to make an impact in the industry, the industry appears to be turning to machines for the much-desired assistance. In fact, the use of artificial intelligence to spot unusual patterns and inconsistencies has rapidly become common for insurance entities. The technology is usually used identifying rings of fraudsters or even in solving auto accident cases.
According to Jim Guszcza, Deloitte Consulting’s chief data scientist in the US, artificial intelligence is part of modern insurance and companies cannot afford not rely on it for assistance. Predictive analytics and machine learning are essential as far as insurance claims are concerned.
Lemonade is among the leading insurance companies that are leveraging the power of data. The New York-based insurance startup was established back in 2015 by two tech gurus. According to the company’s CEO Daniel Schreiber, their data-powered technique allows Lemonade to assess and pay out claims a lot more quickly than most traditional insurers. He added that claims could be accepted and paid out instantly upon approval by Lemonade’s algorithms in nearly one-third of all cases.
Despite the involvement of machines, humans can still go through claims even after they have been paid off. Doing so allows them to train algorithms what they need to be suspicious of in a given claim in a similar way as the machines can spot suspicious factors that they could miss. Daniel Schreiber added that Lemonade is currently finding that its claims unit is responding more quickly because of the existing competition with an algorithm.
According to James Quiggle, the Coalition Against Insurance Fraud’s director of communications, a portion of the zeitgeist amongst insurers currently is no-touch or low-touch claims processing. He also added that more insurance companies are seeking help from machines to help them in dealing with common basic scams. In turn, the move has freed up investigators to cater to more sophisticated matters of their investigations, which can only be dealt with by humans.
Hanzo, a US and UK based web analysis and archiving company, creates software that insurers can utilize in pulling and sifting through volumes of data from marketplace sites, social media platforms or even other web sources in researching claims. According to Keith Laska, the firm’s chief commercial officer, they can effectively collect anything you that you find in a browser. Hanzo’s web crawling tech then sifts through the thousands of content pages to identify the appropriate information.
With the growing use of disparate data sources and machine learning, experts are concerned that the trends may trigger privacy questions. In turn, this situation could lead to the formulation of industry regulation or standards.
