Insurers’ appreciation for orthogonal data

Orthogonal data

It is anticipated that within the next three years, on average every human being on the planet will create about 1.7 megabytes of new information every second. This includes 40,000 Google searches every second, 31 million Facebook messages every minute, and over 400,000 hours of new YouTube videos every day.

At first glance, the importance of this data may not be obvious. But for the insurance industry, tapping into this and other kinds of orthogonal (statistically independent) data is key to finding new ways to create value.

A clearer picture of individual risk

By paying closer attention to the data people create as part of their everyday lives, insurance companies can better anticipate needs, personalize offers, tailor customer experience and streamline claims. Using a wider variety of information is especially useful in better understanding and managing individual risks. For instance, behavior data from sensors, shared through an opt-in customer engagement program, provides insurers with the insight needed to initially assess and price the risk, and mitigate or even prevent subsequent losses.

Take, for example, the use of telematics data from sensors embedded in cars and smartphones. When shared, the raw telemetry data provides insurers with insight into an individual’s actual driving behaviors and patterns. Insurers can reward lower-risk drivers with discounts or rebates while providing education and real-time feedback to help improve the risk profile of higher-risk drivers. Geofencing and other location-based services can further enhance day-to-day customer engagement. In the event of an accident, that same sensor data can be used to initiate an automated FNOL (first notice of loss), initially assess vehicle damage, and digitally recreate and visualize events before, during and after the crash.

Using individual driver behavior to monitor and manage risk is just one way to leverage orthogonal data in insurance. Ultimately, new behavioral and lifestyle data sources have the potential to transform every aspect of the insurance value chain. Forward-looking insurers will tap into these emerging data sources to drive product innovation, deepen customer engagement, improve safety and well-being and even prevent insured losses. For those who invest in the platforms and tools needed to harness the value of orthogonal data, the advantages will be significant.

Reasons why insurers need AI to combat fraud ahead of time.

AI to combat fraud

The insurance industry consists of more than 7,000 companies that collect more than $1 trillion in premiums annually, providing fraudsters with huge opportunities to commit fraud using a growing number of schemes. Fraudsters are successful too often. According to FBI statistics, the total cost of non-health insurance fraud is estimated at more than $40 billion a year.

Fighting fraud is like aiming at a constantly moving target, since criminals constantly hone and change their strategies. As insurers offer customers additional ways to submit information, fraudsters find a way to exploit new channels, and detecting issues is increasingly challenging because threats and attacks are growing in sophistication. For example, organized crime has found a way to roboclaim insurers that set up electronic claims capabilities.

Advanced technologies such as artificial intelligence (AI) can help insurers keep one step ahead of perpetrators. IBM Watson, for instance, helps insurers fight fraud by learning from and adapting to changing business rules and emerging nefarious activities. Watson can learn on the fly, so insurers don’t have to program in changes to sufficiently protect against evolving fraud at all times.

insurers need Artificial Intelligence to combat fraud

Here are four compelling reasons insurers need to begin to address fraud with sophisticated AI systems and machine learning that can continuously monitor claims for fraud potential:

  1. The aging workforce. There are many claims folks who are aging out and will soon retire, taking years of knowledge with them. Seasoned adjusters often rely on their gut instinct to detect fraud, knowing which claims just don’t seem right, based on years of experience. However, incoming claims staff don’t have the experience to know when a claim seems suspicious. Insurers need to seize and convert that knowledge, getting it into a software program or an AI program so that the technology can capture the experience.
  2. Evolving fraud events and tactics. Even though claims people may have looked at fraud the same way for years, the environment surrounding claims is always changing, enabling new ways to commit fraud. Fraud detection tactics that may have worked 6 months ago might not be relevant today. For instance, several years ago when gas prices were through the roof, SUVs were reported stolen at an alarming rate. They weren’t really stolen however — they had just become too costly to operate. Now that gas prices have gone down, this fraud isn’t happening as often. If an insurer programs an expensive rule into the system, 6 months later economic factors may change and that problem may not be an issue anymore.
  3. Digital transformation. Insurers are all striving to go digital and electronic. As they make claims reporting easier, more people are reporting claims electronically, stressing the systems. At the same time, claims staffing levels remain constant, so the same number of workers now have to detect fraud in a much higher claims volume.
  4. Fighting fraud is not the claim handlers’ core job responsibility. The claim adjuster’s job is to adjudicate a claim, get it settled and make the customer happy. Finding fraud puts adjusters in an adversarial situation. Some are uncomfortable with looking for fraud because they don’t like conflict. A system that detects fraud enables adjusters to focus on their areas of expertise.

In the past, insurance organizations relied heavily on their experienced claims adjusters to identify potentially fraudulent claims. But since fraudsters are turning to technology to commit crimes against insurance companies, carriers need to turn to technology to help fight them. Humans will still be a critical component of any fraud detection strategy, however. Today, insurance organizations need a collaborative human-machine approach, since they can’t successfully fight fraud with just one tactic or one system. To fight fraud, humans need machines, and machines need human intervention

error: Content is protected !!