Social media has quickly become the foremost activity on theInternet. The explosive growth in user-generated content has been aboon for insurance claims adjusters and fraud investigators. Navigating the social media landscape,however, can be tricky.

When the World Wide Web was established, early websites weremostly electronic versions of documents. They did a good job atbroadcasting information to a wide audience but were not effectiveat brokering communication. With the advent of Web 2.0 and socialmedia platforms such as Facebook, YouTube, and Twitter, bidirectional communication became one ofthe key reasons to use the Web.

The adoption rate for the use of social media in our daily livesis astounding. When radio was introduced, it took 38 years to reach50 million people. Television was faster, reaching 50 million inonly 13 years. The iPod reached this milestone in three years. Facebook did it inmere months. Social media technology is changing the way the worldcommunicates. Half of all mobile Internet traffic in the U.K. isfor Facebook. It should be pointed out that it's not just kidsusing social media sites. The Pew Internet & American Life Project reports that 46percent of all American adults use at least one social networkingsite.

There are literally hundreds of social media sites on the Webtoday offering platforms for sharing all kinds of information,including photos, videos, status updates, and location check-ins.This has resulted in an explosion of user-generated content. Today,we create as much information in two days as we did from thebeginning of time through 2003.

Given the swell in user-generated content housed on socialmedia sites, it is no surprise that investigators turn to them whenconducting background investigations.

Doing the Digging
With all of this potentially useful information hiding in socialmedia sites, many investigators are adding online social mediaresearch to their investigative strategies. The SIU director of a major insurer says, "depending on the type ofcase, it could be a routine investigative action."

Social media provides a form of voluntary (and free)surveillance. As users create new content, the data they produceand the metadata stored by the platforms they use can be atremendous resource for claims investigators. Information can befound on claimants, witnesses, vendors, medical providers, and evenloss events themselves. The Coalition Against Insurance Fraud (CAIF) reports several caseswhere social media sites helped SIU investigators, including a casewhere "an injured worker who claimed he was too hurt to get out ofbed, yet he had time to update his Myspace page and post upcomingdates and venues his rock band was playing."

The use of social media information during fraud investigationsis not limited to injury claims but can also be used across alllines of business. Let's look at a few examples:

• Auto. While reviewing a suspicious accident involvingtwo drivers that allegedly do not know each other, investigatorsdiscover they are in fact "friends" on Facebook. This suggests thatthe accident may have been staged.
• Property. Upon receipt of a homeowner's fire loss, investigators discover amateur videofootage of the fire in progress on YouTube. This footage clearlycontradicts statements made by the insured.
• Workers' compensation. An injured worker's LinkedInprofile lists a new employer during the time she is continuing tocollect benefits for temporary total disability.
• Liability. A claim against an insuredrestaurant is filed when a customer complains that he discoveredglass in his food. A review on Yelp written by another customerdescribes the event and mentions that the claimant "was obnoxiousfrom the moment he sat down" and appeared to be "totally fakingit."

Following the Rules
While social media sites can provide great information forinvestigators, this capability comes with great responsibility.Even though there is a growing collection of case law on thesubject, the rules and regulations are not clear. Investigatorsmust be careful when conducting social media research since itsmostly uncharted legal waters.

In November 2008, Lori Drew was convicted of computer fraud whenshe created a fictitious Myspace profile in order to taunt one of her daughter'sclassmates, who committed suicide after being subjected to Drew'scyberbullying. Though the conviction was ultimately overturned,the case serves as a warning to investigators. Regarding the Drewcase, investigative researcher Tamara Thompson writes in her blog,"Let this be a warning to information researchers and investigatorswho are collecting data on subjects through their social networkingsites … keep in mind: Is this legal and ethical?" For this reason,it is important for a company to establish strict rules about theproper use of social media sites for investigative purposes. OneSIU director indicated, "The potential for someone to beoverzealous in their attempts to investigate a case using socialnetworking sites is significant … so getting permission to accessthem comes with a pretty high responsibility." While it isacceptable to passively gather publicly available information fromthese sites, investigators must comply with the user agreements foreach social media service and avoid the use of pretext.

Accessing Protected Information
Most social media platforms offer some form of privacy protection to users. In many cases, the user can choosefrom a variety of options for sharing or protecting content. Insome instances, the settings are universal—all content isprotected. Other times, users may be able to choose a privacysetting for an individual post, photo, or other content. One thingis clear, however. Information posted on these sites is intended tobe shared with someone. This is important because courtshave held that this reduces the expectation of privacy and has madeit somewhat easier for insurers and attorneys to access private orprotected information on social media sites. 

While investigators are free to view publicly accessibleinformation, the greatest level of detail is obtained when a useraccount can be accessed directly. Doing so requires some additionalsteps. If a claim is in litigation, for instance, then defensecounsel may issue a subpoena to the social media site or attempt tosecure the data via interrogatory.

More courts are allowing access to such information, especiallyif the publicly accessible content suggests the information may befavorable to the other party or insurer. For example, in Romanov. Steelcase Inc., a 2010 New York Supreme Court case, Romanosued Steelcase because of "permanent injuries" that prevented herfrom leading an active lifestyle. Steelcase argued the publicportions of Romano's Myspace and Facebook pages showed she wasliving a more active lifestyle and that Romano should be requiredto turn over all social media content, even the information thatshe had only intended to share with her friends.

The court ruled that precluding Steelcase fromaccessing Romano's profiles "would condone [her] attempt to hiderelevant information behind self-regulated privacy settings." Itwas reasonable to conclude that the profiles "may contain furtherevidence such as information with regard to her activities andenjoyment of life, all of which are material and relevant to thedefense of this action."

The New Frontier
While many SIUs have added social media research to theirinvestigative toolbox, it is often overwhelming to deal with themassive amount of social media content. Many insurers arestruggling to adequately train analysts and investigators in thelatest search techniques, trying to find the balance between usinga tremendous (free) information source with conducting efficientresearch. Today's research is largely a manual exercise, scouringthe social media landscape looking for useful information about aclaimant and a given loss event. It will not be that way forlong.

Advanced analytics will change the way social media is used ininvestigations. Powerful technologies like text mining, sentimentanalysis, content categorization and social network analysis canautomate social media research and be integrated into the fraudidentification and predictive modeling process. In the future,fraud detection engines will be based not just on claims and policydata, but also on Facebook friends and Twitter posts.

A New Resource
While there are cases where individuals essentially confess tocommitting insurance fraud on their social media profiles—usually in theform of bragging to impress their friends—it is rare to find asmoking gun. Instead, social media serves as a useful, open-sourceintelligence research tool to identify leads. It is critical tocombine social media research with traditional investigativesources for a more complete picture. Certainly people can postmisinformation or exaggerations, as well. Therefore it is importantto validate and corroborate the information discovered online withother sources whenever possible.

Social media sites can be a tremendous resource, butinvestigators must proceed with caution. Social media sleuthingshould always be an integrated part of a larger investigation usinga variety of tools to make a prudent claim decision. It definitelyprovides an advantage to the investigator resulting in valuableleads for further probing.