In 2022, Beneva Inc.’s fraud management department noticed unusual activity under a plan member’s account.
The insurer’s artificial intelligence tool, used to analyze millions of benefits claims, flagged that a plan member and the person’s dependant had used up all their maximum coverages within a short period of time. When the plan coverage rolled over at the beginning of the next year, they also maxed out their coverage in the same timeframe.
“We have a sizeable number of participant plan members, plus their dependants, so . . . AI allows us to prioritize certain cases,” says Esther Gadoua, senior director of business audits, fraud management and dispute resolution for group insurance at Beneva. This was a case that normally would require a huge investigative team to unravel, but in this instance, an AI tool helped flag the abnormal pattern, allowing the company to uncover the fraudulent usage in very little time.
AI tools allow insurers to pre-analyze data gleaned through algorithms and prioritize which data points need to be reviewed for potential cases of benefits fraud. Its ability to process high numbers of data points has allowed the insurance industry to more efficiently weed out benefits fraud. “The amount of data to consult can be very costly before pinpointing a situation that requires further investigation,” says Gadoua. “We have a responsibility to [execute proper] administration of our groups and plans and to . . . manage our costs.”
As the insurance industry becomes more and more digitized, AI will take a more prominent role in flagging abnormalities, irregularities and abuse or fraud. Insurers are constantly having to evolve in all fraud spaces because of the rapid digitization of many industries, including the benefits space, says Vivianna Botticelli, vice-president and chief audit executive at ABC Benefits Corp.
Indeed, benefits fraud is becoming increasingly complex due to the amount of information available, so there’s a greater need for solutions that can analyze these massive amounts of data associated with claims submissions, says Jason Fontaine, a supervisor at Claim Watch, GreenShield’s fraud department.
Alberta Blue Cross, Beneva and GreenShield are among a group of insurers that recently joined an industry initiative launched in 2022 by the Canadian Life and Health Insurance Association to pool claims data and use advanced AI tools to enhance the detection and investigation of benefits fraud. The tool analyzes industry-wide anonymized claims data and identifies patterns across millions of provider records.
“Data is very, very crucial . . . so being able to have [access to] vast data sets [from] different insurers really helped in identifying those abnormal behaviours better,” says Botticelli, noting it’s enabled the insurer to determine when there is a basis for a joint investigation.
“The data set tells a story. It’s very similar to a data fusion centre that policing operations use. And it’s . . . evolving in the insurance space.”
While there’s always going to be a need for the human touch to interpret the fine details of the data, Fontaine believes the use of AI in benefits fraud detection will become much more vital, particularly as cross-collaboration between insurers becomes more standard practice.
To remain competitive, insurers need to use these tools to capture trends, the distribution behind the data and to make decisions and predict, with some accuracy, potential instances of risky behaviour, says Nazanin Tahmasebi, manager of analytics and reporting at Alberta Blue Cross.
She says the results her team has gleaned from using AI have informed decisions on plan design, internal controls and educational campaigns. Since the tool was implemented, Tahmasebi says it’s catching cases of fraud before they escalate and materialize at a significant dollar amount. “That means we’re catching it way faster.”