Tag Archives: john standish

Enhancing Claims Experience With AI

Insurtech and artificial intelligence (AI) have become the new buzz words and mantra in the insurance industry. Creativity and innovation are thriving in Silicon Valley with more than 1,600 technology companies in the insurtech space for underwriting and claims. If you remember, back in the 1990s, experts predicted that if your company was not an internet company, you would not be around for long. That prediction came true, but what about the current prediction that artificial intelligence for claims will change the insurance industry? I believe that it will, and now is the time to
stake your claim for the future.

For the last few decades, insurance companies have yielded massive amounts of data. We know that a robust AI system for insurance claims needs a lot of data to thrive and perform. Because you have the data, how can you start to leverage it and improve your business performance and outcomes? Insurance is now ready for an analytics platform such as Infinilytics’ smartCTM that can help minimize claim costs, and improve customers’ claims experience.

See also: Strategist’s Guide to Artificial Intelligence  

One of the most contemporary AI solutions for claims is fraud and litigation prediction. Imagine the capability of predicting with a 90% accuracy rate if a claim is going to fall into litigation. Your claims team would be able to take immediate steps to mitigate the litigation, customer service would be greatly enhanced and your
claims costs would be reduced.

Insurance companies need to embed A.I. solutions along with their human intelligence so there’s an effective feedback mechanism. Such AI-based claims SaaS solutions aim to:

  • Lower loss-adjustment expenses
  • Reduce the time to settle a claim
  • Identify suspected fraudulent claims with AI pattern matching
  • Use sentiment and emotion analysis for litigation predictions

AI solutions using machine learning require careful deployment and breathing time to achieve the return on investment. Machine learning can cause data biases and hide the context of predictions, which makes it unusable by claims organization. Claims organization need to become super users of such AI solutions and understand the context of these predictions and continue to contribute through continuous feedback. Most successful companies have combined human intelligence with AI to change labor-intensive, pattern-driven processes such as claims.

See also: Insurtech: Can It Help Claims Experience?  

Insurance companies have a vast amount of data. We need to leverage the data and transform the claims process into an efficient and cost-effective business model by quickly bringing artificial intelligence into the claims process.

2 Concepts on Social Media and Analytics

Recently, our team attended the Silicon Valley Insurance Disruption symposium and were privileged to engage in a number of conversations with insurance industry leaders, and innovators in technology. One of the conversations I had with an individual revolved around social media and how it has changed our society; the way people communicate; and how people demand instant information and results. Think about digital disruption in marketing and selling insurance products, i.e., the millennial (on-demand) generation.

The term “social media” means a lot of different things in the insurance industry. I like the concept of using social media in two distinct methodologies; one, strategically; and two, tactically for the business of insurance. For strategic purposes, your company can use social media to communicate to customers and potential buyers of your products with a concise and cogent message. In addition, you can monitor and respond to what customers are saying about your company. For the tactical application, social media can be used for the underwriting and claims business processes. For example, tracking catastrophic events, like significant earthquakes, fire storms and hurricanes; monitoring what people are communicating while these events are occurring is invaluable for your response planning. For risk management in underwriting, the proper information can be obtained in a robust social media analytics solution that can be leveraged for better decision making.

See Also: Does Social Media Have a Place?

The key is to use next-generation analytics with dynamic modeling.

Privacy and the ability to obtain information from social media websites have been analyzed and debated by governments and business leaders around the globe. Actions have been taken to protect an individual’s privacy. So how do you correctly utilize social media analytics in your business process?

Dynamic modeling!strat

Dynamic modeling of social media data begins with identifying the proper source of information. Dynamic modeling provides answers to questions in the underwriting process and claims process while keeping in mind the insured’s interest and experience. Is the API (application programming interface) open to web crawls? Can you filter out the vast amount of noise in the data? (An open-end Google-type search is not the best way to begin). Can you control the overwhelming amount of information? How do you know what to look for? If your team knows the “why” and the actual focus of deterring the risk, you will be able to deliver to the customer a better product, with a competitive price in the marketplace.

Now is the time to start moving forward with next-generation analytics and start being innovative in the marketplace.

Speed To Detection: A Progressive And Strategic Concept Using Advanced Anti-Fraud Analytics

The recent natural disasters in Oklahoma and New Jersey, and the wildfire season in the western United States, have a lot in common when one thinks both of insurance risk — plus the intended and unintended consequences of these events.

The insurance industry knows natural disasters will happen. The industry thus creates and follows protocols and response plans. For the most part, the industry and public-safety officials handle the crisis, and restore calm and order in our communities.

The insurance industry knows these events will occur, and planning is generally pretty solid per the axiom, “If it’s predictable, it’s preventable.”

But in the world of insurance fraud, many sectors of the insurance industry seem to lack the same energy to mitigate this crime. Using the same acumen gained from restoring order after disasters, the key is to apply the same proven strategies of history, response, performance and mitigation of future risks. This approach will better help combat insurance fraud with equal success.

The modern strategy of “speed to detection” is a uniting principal and operating strategy for mitigating the epidemic of fraudulent claims.

Optimizing speed to detection involves synchronizing all layers of insurer personnel into informed, enterprise-wide fraud fighters. They are well-trained to spot warning signs of this crime, personally motivated, and encouraged to follow internal processes that allow open lines of communication about fraud leads, needed process improvements and action solutions.

Bogus claims thus can be discovered and mitigated faster. Quick detection also is an intimidating deterrent that can convince more fraudsters to avoid trying to breach that insurer. The risk of arrest and conviction is too high, and odds of financial reward are too low.

Speed to detection is a timely precept: Insurers today are confronting a persistent crime that is morphing, in many respects, to higher levels of sophistication and ability to steal insurance money.

Insurance fraud harms law-abiding consumers (higher premiums), aids the underground economy, facilitates other illegal enterprises such as trade-based money-laundering, and poses a public-safety threat (e.g., staged automobile collisions, arson, murder for life insurance, needless medical procedures).

Conservatively, fraud steals $80 billion a year across all lines of insurance.1 Some estimates rate the annual losses much higher.

And the problem is growing. Questionable property-casualty claims in the U.S. have increased 27 percent in 2012 over 2010, the National Insurance Crime Bureau (NICB) says in an analysis of its database of claims released in May.

That reflects 91,652 questionable claims in 2010 compared to 116,171 claims in 2012.2 Similarly, most consumer research reveals a disturbing public cynicism about this crime, and even a backslide toward higher consumer tolerance of fraud.3

Confronting this epidemic is a large network of organizations dedicated to minimizing fraud as a virulent national threat.

Insurance companies have teams of experts (the Special Investigation Unit, or SIU) trained to deal with suspicious claims.

State law-enforcement agencies have created specialized departments and bureaus dedicated to thwarting this crime.

State insurance departments have strengthened their processes for identifying, investigating and reporting suspicious claims for potential prosecution.

States also have enacted numerous fraud laws and regulations that further strengthen enforcement. More are being added or bolstered every year.

At first glance, these processes appear sound, prudent and presumably effective. A lot of money, personnel and effort have been thrown at insurance fraud. Shouldn’t schemes be going down instead of up? Or at minimum, leveling off?

Many of the following observations are guided by my 32 years of combating insurance fraud, including several years as a Bureau Chief, and one year as the Division Chief with the nation’s largest anti-fraud unit, the California Department of Insurance, Fraud Division. Some academic backup also is cited for added information.

Despite the large defense shield, growing numbers of insurance executives at the decisionmaking levels — inside and outside the anti-fraud ranks — are frustrated about how fraud persists as a costly national epidemic.

To illustrate: In recent years, I have provided consulting and analysis and review of first-party bad-faith cases involving fraud, the actions of SIUs in a claim or series of claims, and expertise for qui tam civil actions by insurance companies.

In these many interactions with insurance executives, anti-fraud directors and other colleagues throughout the industry, the frustrated question they ask most often about fraud is: “Why do we keep throwing money at a crime that never seems to go away?”

Typically they offer two reasons why fraud remains so vexing and persistent:

“The insurance system invites fraud.” Indeed, our insurance system is one of the best in the world. But the most skillful fraudsters effectively exploit weaknesses when the system is not synchronized and calibrated among partners to create a hardened shield.

“We need the best team to investigate these crimes.” Insurance companies and government entities are constantly working to create an elusive Dream Team for investigations. Key ingredients of team members are passion, creativity, and ability to wade through a series of complex conspiracies either to deny a claim, or have an offender arrested and prosecuted.

Many insurers are frustrated because qualified people with the acumen to investigate fraud are hard to come by. Time after time, when insurance carriers lose bad-faith lawsuits involving the SIU and fraud, some of the common denominators are training, unqualified people and bad leadership decisions.

An important reason fraud appears to keep rising is that insurance companies and regulators are slow to recognize the value and impact of anti-fraud technology leveraged with best business practices.

The anti-fraud community needs to rethink its strategies, and examine ways to identify problems and risks before they become crimes.

Resources should be synchronized to optimize speed to detection.

This requires insurers to have their anti-fraud operations well-aligned with their internal corporate structure, strategies and practices — and with external partners such as state fraud bureaus, law enforcement and NICB.

Reaching this goal must start with an honest discussion about technology and other best practices. A major problem is that too many insurers use outmoded methods of fraud detection. These methods have little impact on modern, sophisticated fraud rings that are a significant source of money outflow.

Meanwhile, insurance fraud is evolving and organized crime increasingly is infiltrating fraud. Such rings have been around for years, but their sheer number and growing sophistication are changing the criminal landscape. Many insurers aren’t equipped to counter this new breed of criminal, especially using indicators.

Recently, I gave a presentation at the Insurance Fraud Management Symposium (IFM). This is the largest annual conference of insurer anti-fraud directors, executives and other personnel.4 The presentation covered a major criminal investigation and prosecution involving a staged accident ring in Southern California.

This case illustrates two frequent insurer vulnerabilities: a) over-reliance on weak fraud indicators that allowed fraudsters to penetrate the insurer’s anti-fraud defenses relatively easily; and b) how vulnerable insurers become when they compromise their business processes by speeding up claims payouts by compromising vigilance.

The leader of this criminal enterprise joined me in the presentation. He was under court order to assist the California Division of Insurance in public education after his conviction.

He related how he ran the operation, who he involved, and how and why he targeted specific insurance companies with bogus injury claims from the setup collisions.

He made a chilling point:

“You will never win the war on fraud.”

He urged insurers to avoid over-reliance on the so-called “indicators” they use to identify fraudulent claims. Indicators are a relatively basic investigative tool. Insurers look for specific actions or behaviors that are red flags of possible fraud during the claims process. With staged accidents, for example, indicators might include flags such as multiple people in both vehicles, expensive treatment at the same clinic, and similar last names to suggest a possible family fraud ring.

This ringleader knew the indicators well, probably better than some claims staff. Thus he could rig his crashes and phony claims to easily avoid being detected by common flags. Just as important, he also relied on inexperienced and untrained claims representatives to give in and pay claims with little scrutiny.

“It is a game of poker: Who is going to bluff the best, and who will stay in the game with a winning hand?” he warned.

In a similarly illustrative case, Greg Foshee was educated, articulate and knew the insurance claims system well. He should have. Foshee was a claims representative for one of the nation’s largest property-casualty insurers. He saw large profit potential when his supervisor ordered him to “just process the claims.”

So Foshee went to the “dark side.” He started staging vehicle accidents and then helped process the ensuing bogus injury claims without insurer scrutiny.

He staged more than 82 vehicle collisions that stole $1 million worth of insurance money. During questioning after his arrest, Foshee said his supervisors told him: “Don’t ask too many questions, just get the claims off your desk.”

Foshee used multiple individuals with multiple valid drivers licenses from several states. He kept the operation simple to avoid detection. He had only 13 ring members, with just three cohorts working full time and controlling the group.

Nor did Foshee involve attorneys and physicians. They would have slowed the claims, and he would have had to split the ill-gotten insurance money with them.

He made smaller claims just for vehicle damage and minor medical treatments in order to stay under insurer radars. The treatments usually consisted of an emergency-room visit for subjective injuries such as whiplash that are typically associated with minor traffic accidents.

Foshee also knew that if his ring members went to emergency rooms too often in a given city, someone might notice and start asking questions. So instead he created false medical bills and treatment reports using letterhead and forms stolen from the hospital.

If the targeted insurance companies had simply called the hospitals to verify patient information, they would have discovered that the so-called patients were never treated there. This would have confirmed that the treatment reports and bills were false.

Foshee averaged $10,000-12,000 income per staged accident, and went undetected for several years. He knew how the claims process worked, and how to avoid scrutiny and detection. The California Highway Patrol’s Investigations Unit completed the investigation in 1988. Foshee was convicted of insurance fraud, conspiracy, grand theft, and was sentenced to several years in state prison.

Let’s think about this for a minute … These aren’t isolated cases. Over the last 30 years, large segments of the insurance industry, law enforcement and other government agencies have relied heavily on old-fashioned indicators of false claims, and similar basic tools. These indicators have been identified, written, promulgated, and used in the daily business of receiving and closing insurance claims.

Reality check, please?

The crime rings knew the insurers’ fraud indicators, and avoided them. The insurers also compromised their internal anti-fraud processes to turn around claims quickly. Many other organized fraud groups and bold criminal entrepreneurs like these are operating daily, skillfully compromising the insurer claims systems. Collectively, they likely steal millions of dollars everyday. Whether detected or undetected, usually it is too late to recoup the stolen money.

Rethinking The Fraud Fight
If speed to detection is to move from an energizing concept to transformative anti-fraud practice, fraud fighters must step out of the indicator box and rethink their entire approach to combating modern, emerging threats such as complex and organized crime rings.

Some insurers just seem to be going through the motions of fighting fraud, indicators and all. But the more progressive insurers are transforming their internal cultures and business practices to create a coordinated, enterprise-wide response to this crime.

They are taking the fight more directly to the criminal underworld instead of waiting for the underworld to come to them. As a result, these insurers are also far more resistant to schemers of all kinds.

Insurance companies and government agencies need the ability to change direction quickly to address emerging fraud schemes, trends and problems. Nimbleness is a key attribute of sophisticated fraudsters. It also should be a core trait of every insurer’s speed-to-detection process.

The goal is not to eliminate fraud indicators or other basic tools. These tools may play a role in the overall mix of anti-fraud business processes and strategies each insurer custom fits for its own anti-fraud challenges.

Several strategic best practices can help optimize speed to detection.

Advanced Analytics
Advanced analytics rank among today’s most transformative best practices for increasing speed to detection and allowing better-informed decision making.5

Analytics involves the discovery and practical use of meaningful patterns of anti-fraud data. Properly marshaled, advanced analytics can quickly move insurers miles beyond indicators. Analytics can reduce the ineffective pay-and-chase mindset of many insurer detection processes. Analytics also can put insurers quickly on the offensive, and thus dramatically increasing speed to detection.

Advanced analytics tools come in many flavors. Each organization must customize an analytics strategy to its unique challenges. Rarely is there one off-the-shelf software solution. Analytics solutions increasingly are being adopted by some insurers. Among the solutions that are gaining momentum:

Predictive analytics. Allows insurers to uncover suspicious activity in close to real time, and even to forecast the likelihood of potentially fraudulent behaviors.

Text analysis. Insurers can ferret out previously inaccessible data such as an adjuster’s field notes — even handwritten notes.

Social network (link) analysis. Helps an insurer examine relationships among organizations, people and transactions to discover suspiciously related claims that appear unrelated on the surface.

Social media analysis. More insurers recently have begun mining social media for clues. A workers compensation insurer, for example, might uncover a supposedly disabled worker posting photos of his Hawaiian surfing vacation on his Facebook page.

But analytics alone — whether advanced or more basic — cannot reverse the tide of fraud. Analytics must be supported by other best practices and processes.

Some insurers and smaller regulatory agencies believe the cost of advanced analytics platforms is too high, or that they do not have the data to support such robust systems.

But analytics can be affordable by starting small (don’t try to boil the ocean), and strategically planning to gradually layer in advanced analytics into the business process and technology platform. Start small, and build upon the new platform incrementally, first addressing immediate business needs and problems.

Marshall Big Data
Mobilizing big data is gaining wider attention in anti-fraud circles. Insurers are sitting on troves of data, hard and soft. Much is never accessed for fraud-fighting. Insurers can dramatically increase their anti-fraud assertiveness by insightfully accessing, analyzing and mobilizing their large volumes of untapped data.

But the terabytes and even petabytes can overwhelm an insurer’s analytical capabilities.

Insurers must invest in analytic expertise to retrieve, filter and use big data properly. Insurers also must know what questions to ask when mining for big data. This information will be more focused and useful, and avoid the confusion and fuzzy results that too much data can impose.

Limit Pay And Chase
Insurers must re-evaluate their reliance on the ineffective “pay-and-chase” model that drives the anti-fraud-strategies of so many insurers. Using this model, insurers routinely pay claims and then investigate afterward.

But the money is gone by then, and the trail is growing cold. It is rare for an insurance company, self-insured or government program to recover much or any stolen money. In fact, usually no money is recovered.

This is especially true of the larger, complex fraud rings that often operate internationally. They are adept at trade-based laundering of stolen insurance money through shell corporations.

Some insurance rings are learning from criminal brethren such as drug cartels in Mexico and South America. They are effectively laundering stolen money (e.g., proceeds from human trafficking, firearms and narcotics). They wash the money through sophisticated shell companies and corporations involved in global commerce. The money is difficult, if not impossible, to trace and recover.

In the public sector, Medicare once was the poster child for ineffective pay-and-chase practices. But the federal health program for seniors is replacing that approach in part by installing predictive analytics to uncover more false claims before payment.

Take On Difficult Cases
Simply going after safe, low-level frauds (i.e., low-hanging fruit such as an inflated claim from a home burglary) might look good on the anti-fraud unit’s statistics reports.

But this also may ignore the largest fraud problems and sources of claims-money outflow such as modern rings that steal safely and efficiently.

They often are organized like a classic cell network. Ring members do not know each other, nor do they know all activities in the enterprise. But advanced analytics can expose these complex groups and their crimes much faster and more efficiently.

Insurers must commit to taking on the difficult higher-dollar cases such as those perpetrated by organized crime rings, even if it entails considerable cost and personnel. This is essential to diminishing what for many insurers is a significant source of false claims payouts.

Collaboration
Better collaboration is essential to turning the corner on America’s fraud epidemic. This collaboration must include all stakeholder organizations and personnel.

Internal. Collaboration within an organization should be an enterprise-wide endeavor and operational commitment. For example, a) agents and brokers must speak with the claims staff; b) claims staff must communicate with the SIU team about suspicious claims; and c) employees at all levels must be encouraged to speak up and identify vulnerabilities, process breakdowns and needed solutions.

To underscore this point, visit another statement the fraud-ring member said at the IFM conference:

“We know when the insurance company will pay based on the actions and interaction with an inexperienced, and not properly trained, claims representative. And we also know which companies pay claims easily.”

External. Insurers must retain open lines of communication with state fraud bureaus, local law enforcement, state attorneys general, the FBI and other stakeholders.

Insurers in different lines of insurance also must collaborate. Auto, workers compensation and health insurers, for example, may find synergy by comparing best practices and exchanging case leads that may uncover hidden crimes.

Insurers in the public and private sectors also must better collaborate for the same reasons. Many organized crime rings, for example, defraud numerous insurance programs. A large Armenian crime ring in California, for instance, staged car crashes against auto insurers and also bilked Medicare. If public and private insurance programs share case leads, they can dramatically increase the joint knowledge base needed to more speedily break down that ring.

One promising collaborative effort is the new Fraud Prevention Partnership. It was formally announced last July by HHS Secretary Kathleen Sebelius and U.S. Attorney General Eric Holder.6

Medicare, private health insurers, automobile insurers and others are formalizing closer lines of cooperation. The partnership is building up its operating structure, and partnership members are beginning to share fruitful case leads. It could become a model for collaborative techniques.

The Payoff
Marshaling analytics and big data with current rules and indicators into a seamless and unified anti-fraud effort creates an expansive world of possibilities.

Imagine the ability to search a billion rows of data and derive incisive answers to complex questions in seconds.

Imagine being able to comb through huge numbers of claim files quickly.

Imagine more-quickly linking numerous ring members and entities acting in well-disguised concert. These suspects likely could not be detected with sole or even primary reliance on basic methods such as fraud indicators.

Ultimately, imagine analyzing entire caseloads faster and more completely, thus addressing the largest fraud problems and cost drivers in any of an insurer’s coverage territories.

Conclusion
Insurance companies are not in the anti-fraud business. They are in the business of managing a risk pool, mitigating those risks and returning a fair profit. Government law-enforcement agencies are specifically charged with preventing crime and disorder.

To prevent fraud, all involved organizations must scrutinize their systems with a fresh view and openness to evaluating how to better combat this crime.

Advanced analytics, coupled with sound business practices and preventive measures, will yield better anti-fraud results. For insurance swindlers, speed to detection should mean speed to jail.

1 Coalition Against Insurance Fraud, estimate of annual fraud losses.

2 U.S. Questionable Claims Report, National Insurance Crime Bureau, May 16, 2013.

3 Four Faces of Insurance Fraud, Coalition Against Fraud, 2007; Poor Service Leads to Fraudulent Claims, Accenture consumer survey, 2010.

4 An Insider’s Perspective on Automobile Insurance Fraud — Why It Is So Easy to Steal From Insurance Companies, and What To Do About It. White Paper by SAS, 2013.

5 Competing on Analytics, The New Science of Winning. Thomas H. Davenport and Jeanne Harris,
Harvard Business School Press. 2007.

6 New Anti-Fraud Partnership is a Force Multiplier, news release, Coalition Against Insurance Fraud, July 25, 2012.

An Inside Perspective On Automobile Insurance Fraud, Part 2

This is Part 2 in a two-part series on automobile insurance fraud. Part 1 in the series appears here.

Who Participates In This Type Of Insurance Fraud?

Just about anyone. You'd be surprised. Even people who consider themselves upstanding citizens will get drawn into the business, because they see it as a victimless crime. One of the first cases I investigated involved a college-educated, former Farmers Insurance adjuster from Ohio. One day, he just decided to go to the dark side of the earth and started staging collisions from Ohio to California. He got away with $11 million before we caught him and put him in prison. He had so much activity going on that he carried a briefcase with him, and in that briefcase were 13 valid licenses from Colorado, Ohio and Texas — all valid — along with lots of crib notes from all of his activity. In an unlucky turn of events for the fraudster, he was stopped for speeding one night. As he opened his briefcase to get out a driver's license, that sheaf of crib notes was visible to the highway patrol officer, who reached right over his head and grabbed it. Lesson learned. Keep your crib notes to yourself.

People don't necessarily set out to go into insurance fraud as a career, but it's easy to see the attraction, said Borloff. “When you're first introduced to the people in this business, they say, 'This guy, he's in the insurance business,' and everybody understands he doesn't have an insurance company, he's in a different side of the business. But he's a well-to-do guy, with a house in Beverly Hills, with a car, with everything. And you ask yourself, 'Why can he do it and I can't?' And then you start to learn.”

It starts easily enough. The newcomer becomes the defendant in one staged accident, and voilà, $1,000 just falls into his lap. “It's a big deal for this guy,” said Borloff. “He just calls the insurance company, and that's it, he's a rich man. He wants to do it again. Trust me, he wants to do it again and again.” However, a savvy criminal enterprise will not use this guy again. They will use another guy, then another. So now our newcomer is intrigued to work it from the other side of the fence, to run his own business.

First he needs to recruit lawyers and doctors — and that's surprisingly easy, said Borloff. Where would you find an attorney willing to take such a risk? “Word of mouth or the Yellow Pages,” Borloff quipped. “Any attorney wants more business.” The business is so attractive that it is relatively easy to find professionals who want to be a part of it.

“Think about it. Who is this lawyer guy? He spent three years in law school, and he spent a lot of money. He is out of school now, and he has a lot of debt. What can he get? He can get a job for about $40,000 to $50,000 a year at the most, if he is lucky. So he opens his own office. He has just one secretary, and he's waiting for clients. But there are no clients, because there are a lot of lawyers around with the big names, big firms and so on. He is desperate.

“If you come to this guy and say, 'You know, I can give some business to you,' what will he say? As a lawyer, he is like an innocent girl — he wants seduction. You should not say to him, 'It's a staged accident.' No, you say, 'I can give you some business, but you will have to follow my instructions and pay me 50 percent.' He says, 'Of course,' and he agrees. If you teach him properly, he will properly do his business for you.”

Is it really that easy? Are there attorneys willing to look the other way at the likelihood that their cases are fraudulent? Yes. In a separate case, when I had an attorney in an investigation room and got him to confess, I asked him, “What were you thinking?” This was an aha moment. He said, “It's quite simple — there aren't enough real accidents to go around.” It's that simple.

Besides, real accident cases get complicated and messy, Borloff said. “Attorneys don't like real accidents because there are so many problems. There isn't a cooperative defendant — he says, 'No it didn't happen.' There are too many problems. It's easy for me to work with defendants who say, 'Yes, I hit him, I was not paying attention,' and with attorneys who want money.”

Many of the people involved in these schemes don't quite see it as wrong but rather as a rightful Robin Hood redistribution of monies. “I had one good, white-bread American guy. He was the perfect defendant,” said Borloff. “He did the accident, and then on the way home, he sees somebody not driving the proper way. He gets on the phone, calls the police, and gives them the license plate number. I said, 'You turned him in? You broke the law too.' He said, 'No, that was the insurance company.' Sometimes the insurance company is not seen as the good guys.”

Fraud Ring Leader: “This Is A War That Will Never End”

“If you want to win this war, you can't do it,” said Borloff. “It's like the Cold War between the USA and Soviet Union. There was intelligence and counterintelligence, and it went on for a long, long time.” Insurance companies develop more sophisticated fraud detection and prevention tactics, but the criminal enterprises adapt and become more sophisticated too. An insurance company known to have strong defenses may repel fraudsters, but only for awhile. Pitted against the fraud rings, insurance companies are at a disadvantage, said Borloff.

They lack depth in claims adjusters. “A good, educated adjuster knows how to work — he has some police experience, maybe some counterintelligence experience,” said Borloff. “I ask you, how many adjusters do you have with these credentials? I can tell you 1 percent is a generous calculation.”

Even if you can recognize staged collisions, there's still generally a payout. “You can try to fight it, but it's probably just to reduce the amount you have to pay,” said Borloff. Adjusters can try to bluff — say that the examination of the car indicates there weren't four passengers, or that damages are inconsistent with the accident description — but that doesn't work. The stager knows a lie when he hears it.

The legal system doesn't offer much redress. When a claim looks suspicious, a smart adjuster asks for a deposition and asks smart questions of the plaintiff in front of him. He may do his best, probing for details, hoping to ferret out inconsistencies. “But it's worthless,” said Borloff. “It doesn't matter what the answers are. If the people said back pain the first time, and neck pain the second time, what does it prove? It proves nothing. You can't go to court with this stuff. After the deposition, the adjuster doesn't have a lot of choices. He has to negotiate the price.”

The adjuster can make a lowball offer, but a good defender will push back, knowing the adjuster's supervisor will approve more, and the insurance company doesn't want to go to court. In court, the insurance company can only argue on the facts, but the facts are that the incident did happen, and the plaintiffs do report that they suffered pain and injury from it. Can you prove they didn't? You can't pressure defendants to recant, because they're in bed with the criminal enterprise. They've got the money and face jail time if they confess.

“In small claims court, insurance companies have no chance,” said Borloff. “A claims adjuster can come and represent the company, but if it happened, it happened. If you say you have pain, I have pain, we have a medical bill, and you have to pay toward this. I have a limit of $5,000, and each defendant can ask about this amount of money. And what does this situation get insurance companies? Nothing but a lot of trouble, and more trouble. The adjuster doesn't want this problem. His supervisor doesn't want this problem.” The fraud enterprise wins.

Let's have a reality check here. The insurance industry will never stop this. People are going to try to scheme our insurance system as long as the system we have in place today exists, and ours is one of the best in the world. It takes care of consumers when they get in trouble. But with that, you will always have people thinking of ways to scam the money. Change the system, and they come up with a new scheme. This is just what people do.

How One Investigative Team Won

The fight against insurance fraud may be a never-ending war, but there are still skirmishes to be won. I led a five-year undercover investigation of a large and sophisticated organized crime ring in Southern California — Borloff's, in fact — and won.

We started with 63 suspected fraudulent claims that were on file in the San Diego office of the Department of Insurance. There were quite a few interesting outliers in those claims. For one, everybody was of Russian descent. These were all rear-end collisions, sudden stops. Now, we all know that sudden stop rear-end collision is a quick pay. It's pretty easy to determine at-fault in these collisions — a couple hundred of them happen every day in Southern California. It's not that big a deal.

Then an individual in custody for an unrelated matter came forward and offered information that made these claims look particularly interesting — even named two primary players. A formal task force was established with the equivalent of a joint powers authority between federal, state and municipal law enforcement. An undercover officer was introduced into the San Diego community to try to identify the ring leaders and learn how they were recruiting others into the organization and conducting business.

I didn't want to chase stuffed passengers or street offenders at the lower level. I wanted to go deeper into the organization — to identify the stagers, attorneys and physicians who facilitated those claims. And it worked.

A Collaborative Effort
The undercover operation was a joint effort of the California Department of Insurance's Fraud Division, the FBI, the San Diego Police Department, the Immigration and Naturalization Service (INS), and the National Insurance Crime Bureau (NICB). Undercover agents came from the California Department of Insurance, the San Diego Police Department, the California Department of Health Services, and the Bureau of Narcotics Enforcement.

Seven major insurance companies cooperated by providing pretext policies set up solely for the purposes of the investigation. Only the companies' regional vice presidents of claims and Special Investigations Unit directors knew about the investigation. They agreed to have the claims legitimately paid, so the money could be tracked and there would be no suspicion of law enforcement involvement.

A lot of thought and effort went into this to backstop the identities of the undercover people. If a private investigator or lawyer ran these people, they would show up as true legitimate people with valid Social Security numbers, credit histories, houses, vehicles — it was backstopped to the hilt. You'd have no idea that it was law enforcement.

Working From The Inside
We successfully infiltrated this ring for more than 18 months and were staging collisions in San Diego, Los Angeles and San Francisco. Borloff was identified in the very first staged collision, and then the lead undercover detective partnered with Borloff to stage more collisions and car thefts. Borloff gave instructions to his new business partner about how to get involved in this game. For every car and policy he provided, the undercover officer received $2,500 – a total of $75,000 over the course of the investigation.

Little did Borloff know these cars were coming from the NICB salvage pool, and every move was being recorded. During the five-year period, Borloff was responsible for staging more than 100 automobile collisions, exposing the insurance industry to an estimated $2 million in fraud losses.

Five Years To Success
It was kind of dicey. The [cooperating] insurance companies paid out more than $230,000 in claims into this. However, everybody had the bigger picture in mind — the potential economic loss prevented (PELP), an FBI metric that forecasts the money that would have been lost if the criminals continued their activities unabated.

When it was time to strike, the United States Attorney's office in San Diego handled federal prosecution and the Los Angeles County District Attorney's office handled the state prosecution. This was a seminal case that eliminated the issue of double jeopardy. In the State of California, you can arrest and convict someone in federal court for mail and wire fraud with a scheme of insurance fraud, and then charge them at the state level with insurance fraud, and it's not double jeopardy.

I brought the investigation from cradle to grave. I started as the supervising agent in the San Diego Fraud Division of!ce, and by the time the investigation was over, I was captain of that office. The investigation netted attorneys, physicians and chiropractors along with their office staff, eight cappers and 44 claimants — and effectively shut down one of the largest such rings in Southern California.

The Information Imperative

“The main problem for insurance companies is they don't have enough information,” said Borloff. “When the adjuster discusses the case, he knows nothing. He knows just this is the car, this is the car damage, these are the people. He can check the records — see these people didn't have an accident for two years, three years and think they sound like nice people. But he's still suspicious, he pursues it all the way, and still he gets nothing, because it's still trouble, still problems.”

In order to detect suspicious claim activity, insurance companies need access to transaction information and supporting detail that typically resides in different systems. Transactions viewed in isolation could appear normal, but they might look quite different if you could correlate those transactions across related entities. However, a unified view based on multiple internal data sources is rare. Rarer still is the organization that has augmented that view with external data as well.

As part of an 18-month Command College program of graduate study, I designed a virtual office environment for a law enforcement environment. After graduation, that work led to an assignment to design a fusion center to promote information-sharing among federal, state and local entities. The fusion center I designed was the first in the nation to incorporate law enforcement data and financial data, which led to new discoveries. For example, we threw the system up for a test run, and investigators in the San Francisco Bay Area were playing with it. From working with the data and social networking analysis, we found two chop shops in Dallas that the local police department didn't know about.

Smarter integration and analysis of data will be a strong defense to the growing fraud problem, as well as a way to meet the pressures to achieve more with less. The data being gathered through the claims process is growing bigger and bigger, so you need to start looking at things differently and working smarter, and that means leveraging your data together. A number of statistical approaches can be created to build a solid predictive solution. For instance, when you combine business rules, anomaly detection (finding outliers), and social media analysis, you can identify suspicious claims even if there is no prior claim history.

Getting a handle on the fraud problem is not about processing more cases. It is all about working the right cases to make an impact to reduce fraud. If you do not identify the true cost drivers in fraud — the licensed professionals, administrators, cappers, stagers and other individuals controlling the criminal enterprises — you will never truly reduce the amount of insurance fraud in our communities. But when you apply best practices and analytics together, you create a powerful tool and business model to reduce fraud and provide great ROI for anti-fraud programs.

This series of articles is taken from the SAS white paper of the same name. © 2013, SAS Institute Inc. Used by permission.

An Inside Perspective On Automobile Insurance Fraud, Part 1

This is Part 1 in a two-part series on automobile insurance fraud. Part 2 in the series can be found here.

Introduction

Traffic engineers would love to unblock the clogged arteries of Southern California's freeway system, where rush hour is anything but “rush” — more like gridlock.

But in a land where one's car is one's empire, one's freedom and personal statement, carpooling is a tough sell. The high-occupancy vehicle (HOV) lanes have scant occupancy.

In fact, cars carrying multiple passengers are such a rarity that this scenario alone raises red flags for auto insurance claims adjusters.

Operating under the radar is a fast-growing segment of the so-called “underground economy” — organized criminal enterprises that stage automobile collisions with the intent to defraud insurance companies of medical payments. In some cases, the entire incident is created on paper, with fictitious vehicles and false identities. In other cases, the perpetrators take real vehicles with legitimate insurance policies out to vacant lots or remote fields to crash them and then fill out a counter report. The most compelling cases are the ones where participants intentionally ram vehicles together on city streets — often a rear-end collision in a left turn lane — then dial 911 and wait for police and emergency medical services (EMS) to arrive. This approach triggers a police report and EMS records, which lend an air of legitimacy to the event. It really happened.

Based on instructions from a stager, the driver and two or three passengers — who are known as “stuffed passengers” — report neck and back injuries. The passengers later visit a physician or chiropractor who is in collusion with the criminal ring. The patients sign in and leave without receiving any treatment. If the insurance company balks at paying the specious claim, the claimant enlists the help of an attorney who is also party to the scheme. The attorney is tenacious, willing to go to court, generally able to bluff until the insurance company backs down and settles.

In the process, everybody except the insurance company gets easy money. Property damage to the vehicle is paid to the owner of the vehicle, while multiple players split the proceeds of the settlement for medical payments. In a typical case where the insurance company settles for, say, $6,000, each vehicle occupant might get $1,000, the lawyers and doctors collect their fees, and the enterprise leader retains 50 percent of the professional services fees plus the balance of the claimants' settlement, if any. If the enterprise leader successfully stages dozens of such incidents a month, it's a lucrative business.

This practice exploded in Southern California in the mid-1990s. If you are a Special Investigations Unit investigator, you are dealing with this every day. The average caseload for an adjuster or claims representative might be 150 or 200 a day, depending on the size of the company. At least 25 percent of that is some flavor of fraud. It's either a false claim or an embellishment to it. People are doing it. Even people who think of themselves as law-abiding are doing it, because they don't think of insurance companies as victims. This type of activity is so prevalent that our undercover investigators would hear paramedics on the scene saying, “Okay, which one of you is going to the hospital this time?”

Automobile insurance fraud is such easy money that the business is even creating unlikely bedfellows. For example, in South Central Los Angeles, the Bloods and Crips — gangs that have had an intense and bitter rivalry — are now cooperating with one another in organized insurance fraud, because it's more powerful and profitable to join forces.

Six Steps to a Successful Insurance Scam

Constantin Borloff (not his real name), the former leader of a successful and sophisticated fraud enterprise that operated in San Diego, Los Angeles and San Francisco, shares his top tips for making fraud pay. Having paid his debt to society, the ring leader now tells insurance companies how he was able to steal so much money from them, who does it and why it's so easy.

Go For The Med Pay Money
Borloff would insist that vehicle insurance policies have med pay coverage — coverage for reasonable expenses to treat accident-related bodily injury. Since this coverage follows the vehicle, passengers in a vehicle that has med pay coverage will likely be covered as well. Borloff gave vehicle owners a list of insurance companies who would freely provide these policies.

In theory, claimants are supposed to repay med pay money if they receive a settlement, but that doesn't happen according to Borloff. “For all history, maybe two times the insurance company asked for money back. If you say you don't have money and can't pay it back, they say, 'Okay, don't pay back the money.'”

Find the Inattentive Insurance Companies
Borloff also selected insurance companies with a reputation for laxity, the ones whose claims representatives didn't take a stand and ask the hard questions. “Big companies like State Farm or Farmers have millions of policies, good special investigation units and more experienced adjusters, so that's where you would see more problems. It's better to go to the smaller company or where it's not their main business. These companies usually pay more, while the big companies usually pay a little less.”

Insiders in the business share this information, so they know which companies to avoid and which ones would pay off like loose slot machines in Henderson, Nevada.

What would make an insurance company an unattractive target? “I don't know what will stop me,” said Borloff. “All insurance companies are bound by law to pay. So for us, the system is working perfectly. The insurance company can fight, and they have a lot of resources to fight, but eventually they have to pay something. Maybe more, maybe less, but eventually they have to pay something.”

Choose Participants Who Won't Raise Suspicion
In a perfect world, your participants are white American citizens with clean driving records and their own drivers' licenses. Judges and juries look most kindly upon this type of claimant, according to Borloff.

It is equally important that their behavior fits accepted patterns. For instance, policies would be active for four to eight months before the staged collision. Claims would be modest, usually no more than $5,000 or $6,000. Activities were choreographed to avoid triggering red flags. “I know insurance companies have about 25 red flags,” Borloff says. “What the claims adjusters know, the criminal enterprise knows twice. I knew about all these red flags, and I tried to avoid them.”

Distributing the cases is one way to avoid detection, said Borloff. “If the enterprise will do, say, 20 collisions a month, the claims will go to five different insurance companies, each to a different attorney — 10, 15 or 20 different attorneys — and any given adjuster will have at most two cases to a specific attorney. Will the adjuster be suspicious about it? I don't think so. It's very dif!cult for the insurance company to catch these people in this situation.”

Borloff tells of a fringe case where a woman, working against the advice of her stager, staged four accidents in a single week. She submitted claims to four different insurance agencies. All four claims were paid, but this pattern of activity could have exposed everybody in the fraud enterprise to scrutiny and discovery.

Pay More Than Lip Service To The Medical Treatment
When private investigators were first sent to wait outside medical clinics to observe and videotape (the comings and goings of visitors), the first people they caught were the ones who walked in, signed in and left within a minute. People quickly learned to stay longer inside the clinic and have follow-up visits at intervals that would seem appropriate for their injuries and type of care.

Keep Your Stories Straight
Cappers and stagers write notes for people so they can remember their stories when talking to claims representatives, and later on, if they meet with an attorney and go into depositions. Somehow, somewhere, there is a record of all this. If the ring is dealing in volume, there must be good notes, or they won't remember the details of a case, and that's how they get tripped up. Some stagers get tripped up simply by having these notes in their possession — in their offices or briefcases, waiting to be found during a routine traffic stop or search.

Insulate The Players From Each Other
These groups tend to function as classic cell networks. In an effective cell network, the claimant may or may not be exposed to the other people involved, or may be only exposed to the doctor but not to the attorney. That's how these people are protected from one another. Participants may not have a knowledge of what else the group is doing. When we arrested 72 people on a state level and brought them into interrogation rooms for 72 hours, it was pretty clear that they only knew their own activities or those of friends they had brought into the group. They had no knowledge of the bigger scheme. That's how you protect your enterprise.

The parties in these fraud rings learn never to admit to anybody that the accident was staged. Everybody in the enterprise knows it, but if you tell even one person, there's a point of vulnerability. It is especially important to insulate the medical and legal providers, because their professional licenses are critical to facilitate these claims. They take it all the way and never back down.

How often would a criminal enterprise walk away from a case because an insurance company's Special Investigations Unit got involved? “I would not walk away, but I would accept lower settlement, for sure,” said Borloff. “One time one of my colleagues made a terrible mistake, and sent 63 cases to Allstate — one attorney, same office. They came to me and said, 'What should we do now, SIU is after us?' I said, 'Don't give up, try to fight,' but they decided to give up. It was the biggest red flag. They lost money. It upset people.” Giving up is tantamount to an admission of wrongdoing.

This series of articles is taken from the SAS white paper of the same name. © 2013, SAS Institute Inc. Used by permission.