Insurers will increasingly use public shared data to inform pricing decisions and aid in fraud detection, predict researchers from Celent.

What’s more, says Mike Fitzgerald, a senior analyst in the Chicago office of Celent, insurers could use information gleaned from social networking websites, including Facebook, when deciding whether and how to underwrite applicants for life insurance.

For instance, an underwriter might notice on a social network that a certain applicant has indicated he or she is a “fan” of risky hobbies like skydiving or rock-climbing, or has a circle of friends who are active in these sports. That information might be interesting to the underwriter, he says, especially if the life insurance application indicates that the applicant does not do skydiving.

“The underwriter may decide to ask some follow-on questions about this to the agent or the client,” he says.

In another example, Fitzgerald says, a social network website might indicate that an applicant for life insurance is “very” involved in the American Cancer Society. Such active involvement might cause the underwriter to follow up to see if a close family member, such as the mother or father, has cancer, he says.

The answer could impact the underwriting decision, he says.

The comments were spurred by findings in a just-released report from Celent on how insurers are, or could be, leveraging social networking tools. In that report, two Celent researchers — Catherine Stagg-Macey and Craig Beattie — concede that insurer use of such information in pricing decisions is not without challenges.

“It is difficult to get a confirmed match between a customer’s account on a social networking site and the details the customers share with an insurer,” the two researchers write. So, “an insurer must then ask for the customer’s account name in order to get the match, and this clearly may not be offered if it becomes known that the information could increase the premium.”

Another challenge would be persuading customers to share their social network data and then using it effectively in producing new insurance products and improved pricing, the researchers continue.

Some customers may end up benefiting from preferential rates and better service, they allow, but “customers who associate with poor drivers, criminals and friends involved with fraud may find that they pay a higher premium or are not offered insurance.”

Fitzgerald predicts that insurers will initially seek out social networking data for applicants seeking high face value life policies. “It would be too expensive for underwriters to obtain this information on applicants for smaller policies because the searches are not automated,” he explains.

There are no information companies that currently provide this data to underwriters, but he says they could develop and become “a whole new industry.”

Fitzgerald does not believe that the use of such information will stir up privacy concerns.

People who put personal information on social networks can restrict the degree to which they want this information to be publicly accessible, he says. But some people are “over-sharers,” he says, noting that they make a lot of personal information publicly available.

“Underwriters can–and should have the right to–use this information,” he contends.

“That’s absurd,” says Hank George, and underwriting expert who is president of Hank George Inc., Greendale, Wis.

To obtain personal information from social network websites “is invasive of people’s personal space, and it’s going in the wrong direction for a responsible industry,” he maintains.

“In addition, I question whether the assumptions the underwriters make from such information would be valid. For sure, there would be lawsuits a-plenty, if the life insurance industry goes in this direction.”

For example, George says, what happens if an applicant used to do skydiving many years ago but no longer does so? “If a social network website mentions his interest in the sport, the underwriter might draw inferences from this that are absolutely not true.”

Similar discussions are circulating in underwriting circles about using predictive modeling for life and other mortality-based underwriting, George points out. He says he opposes this too.

“I’m opposed to getting personal information without the express authorization from the customer. We shouldn’t be getting unauthorized information from which we draw inferences and then use it to nitpick and make judgments based on information that is not valid. That’s not fair or equitable to the customer.”

The insurance industry already has access to valid data, George adds, citing court records, motor vehicle reports, licenses and pharmacy profiles as examples.

“For a more detailed look, underwriters can also use teleunderwriting–personal history interviews over the telephone with the applicant or designated representative.

“We don’t need to go prying into someone’s personal space and make inferences….That could put the right to underwrite insurance products in jeopardy. It would be like internet-based redlining–i.e., punishing people by using whatever information the underwriter can get to discriminate unfairly.”