Social media is changing the nature of insurance fraud investigations, as sleuths uncover connections among criminals by digging through popular websites — and find that scammers are often their own worst enemies.
The scam seemed easy enough. A woman worked in the facilities department at a company that awarded a maintenance contract to a firm secretly run by her husband. Soon after, he started billing for all kinds of services — most of which were not actually rendered. The fraud seemed simple enough to execute — and with insiders at both companies, it could prove hard to detect.
“He pads invoices and she pays them,” said Chris Giovino, partner in charge of forensics at Dempsey Partners. “They were outwardly stealing from the company.”
But in this case, the fraudsters didn’t count on Giovino, who spent 30 years as a special agent with the Drug Enforcement Administration. Giovino had the requisite expertise in mining social media and databases for clues about fraud, and his investigation led him all over the Internet, from Facebook to LinkedIn to the White Pages. He eventually found an address associated with the maintenance vendor that matched an address associated with the couple’s grown children. Giovino had hit the bull’s-eye.
Investigations like Giovino’s are becoming more typical, as insurance companies and third-party investigators say that mining social media sites is an essential part of insurance-related fraud research. No longer is social media a peripheral, cutting-edge tool used in investigations. These days, it’s a pillar of the process.
And with insurance fraud accounting for $30 billion, or 10 percent of losses in the property and casualty insurance industries in the United States each year, according to the Insurance Information Institute, investigators need any effective methods they can get their hands on. Adding to the need is that the fraudsters, especially ones attacking company’s cyber security, are using methods that investigators may not have even heard about yet.
Aaron Soline, a manager at the National Insurance Crime Bureau, said that when he started investigating insurance fraud many years ago, he typically had nothing to go on but a name, maybe an address. With the advent of the connections generated by social media, Soline said he can now find photos, employment information, past education, date of birth and even see who people interact with. It all helps paint a more complete picture of a potential fraudster and the crime for which they are being investigated.
Any discussion of social media starts with Facebook. With more than 950 million users, chances are that site alone is likely to yield information about would-be thieves. In fact, investigators are often helped in their quest by the very subjects they are looking for. Even though Facebook recently upgraded its privacy settings after mounting criticism, “a lot of people don’t care about the privacy settings [on Facebook] or don’t know how to use them,” said Soline. “Twitter has privacy [settings] but most people don’t engage those either.”
Private Web page settings hamper investigations, said Kelly Cory, president and CEO at Keystone Investigative Services Inc., but plenty of people couldn’t be bothered with engaging privacy settings. “If they’re blocked, they’re blocked, but you’d be surprised at how many people don’t think of that,” she said.
Facebook’s newly added location services allow users to “check in” at places to tell friends that they are at a bar, restaurant, park or even a different city, and it turns out that that tool is a big help to investigators. Here’s why: A trucker “checks in” at a bar on his Facebook page at 5:43 p.m. on Friday. He later climbs into his rig and minutes later plows into three cars at a strip mall parking lot. Police reports indicate the accident took place at 7:39 p.m., and the trucker claims he was on his way to deliver his load. Both true, but the Facebook entry reveals there’s more, and investigators have a reason to ask if the trucker was hitting the bottle before the accident.
Similarly, if an injured worker’s Facebook page indicates he or she “checked in” at the gym, it’s also probable they were working out, leading investigators to question an injury claim filed by a worker pumping iron at the gym.
Investigators stop short of “friending” the person they are probing. That’s because contacting claimants who are represented by attorneys violates the law, said Cory, and “friending” could certainly be seen as a form of contact.
“You don’t want to get into an area where it’s entrapment,” she said.
Scouring the Internet for clues about fraudulent activity may sound easy. Investigators could just run searches on Facebook, LinkedIn and Twitter, then get more background research using Google or Bing, and call it a day.
But the process can actually be quite cumbersome, experts said.
Dead ends persist, false positives abound. A search for “John Smith” could yield thousands of people across the country. So investigators painstakingly try to connect information about people to corroborate the information. Doing so efficiently is paramount.
“It’s actually really boring and tedious. You have to know when to stop or else you can spend 20 hours going into a dead end,” said Giovino. “That’s what separates the good investigators from the bad ones.”
Perhaps nowhere has social media made more inroads in rooting out insurance scams than in workers’ compensation.
An “injured” worker posts a photo of himself riding a mechanical bull. A Facebook friend unwittingly shares a photo of a “disabled” worker playing basketball in a recreational league game. A claimant on temporary total disability posts a video to YouTube of herself in a dance competition.
Cory recalled a time when she was investigating a person claiming to have a back injury, then found a photo posted to a social media site of the individual participating in a karate class. It’s that kind of evidence that can crack a case — and get an insurance company’s attention. “When you put that picture in the report, the clients eat that up,” she said.
Social media has become a godsend for investigators, and fraudsters are often so inept that they verge on the amateurish: They share information about their crimes and invite capture. “People like to show friends they’re riding a jet ski or show them that they’re out and about and doing things,” said Soline of the NICB, “so they end up posting it.”
Jim Quiggle, director of communications at the Coalition Against Insurance Fraud in Washington, D.C., said workers’ comp insurers seem the most pleased with how social media is helping them find fraud.
“People love to brag about how they ripped off insurance companies. They can’t keep their mouths shut,” he said. “People post photos of themselves in Aspen on a black diamond slope” when they should be home nursing an injury.
Brian Smidt, vice president of data analytics at the NICB, said swindlers of all types get caught because they share too much on social media sites. A claimant who reports his motorcycle stolen in March, but then posts a photo of himself riding it in May, practically solves his own crime as investigators can use technical clues to figure out when online photos were taken.
Smidt said there have been many times when his team has presented an individual with evidence of claims fraud gathered on social media sites, and the claimant has subsequently retracted his or her claim. Those cases are easy, as they practically solve themselves, he said.
Investigators find social media particularly useful to track the connections between fraudsters, which can tip law enforcement off to larger, more sophisticated crime rings. “It’s unlikely you’d crash into a car full of people you are friends with on Facebook, when you previously admit that you don’t know them,” said Soline.
Tracking down larger, more organized crime rings is much more satisfying, experts said. If a doctor and an attorney have been working together on fraudulent auto accident cases, are they connected through Facebook or LinkedIn? Are they using Twitter to set up dinner plans?
“It helps locate different relationships you might not be aware of,” said Soline.
Al Marrazzo, manager of the special investigations unit at ESIS Inc., agrees. If a person has a slip-and-fall at a store — then a series of searches find that they live in a house with two people who also have pending slip-and-fall cases — that’s a suspicious coincidence. “Nobody slips and falls that many times,” he said. “There is so much more information at our desktops that we didn’t have in the past — who they are, who they associate with and links through the Internet to other names.”
ESIS, the third-party administrator that manages claims for the ACE Group, hires private firms to investigate fraud — and they are required to conduct Internet and social media website searches as part of their investigations.
That can mean big money savings.
An investigation of an injured worker may lead to an adult basketball league website, and provide the date, time and location of a game. Surveillance teams show up at the game — meaning they only need to be paid for that one occurrence, not days and days of surveillance that might turn up nothing.
Brian Wilson, head of the Special Investigations Unit at Zurich North America, said social media has been a great tool in helping its investigators confirm identities of individuals since they are able to find photos, vehicle information, location, manner of dress or interests. That can help Zurich investigators — who are often former law enforcement officials and claims professionals — create a timeline of the fraud.
But Wilson is careful not to take information on the Internet at face value. Instead, he uses it to corroborate or dispute information from traditional investigative methods such as interviews or records checks.
“There is no formula” for investigations at Zurich, he said. Investigators start with what they have and move quickly, especially because information on the Internet has a “shelf life,” and needs to be discovered before it’s taken down. Auctions on eBay eventually come to an end and are removed from the site. Newspaper stories get archived. People take down YouTube videos. Investigators need to work fast to make sure they get all the information before it disappears.
(From Risk & Insurance, October 2012)