Extreme weather is tied to insurance fraud. How can insurers minimize costs and still serve legitimate policyholders?
Hurricane season has arrived in North America, but insurers should be thinking about extreme weather year-round. High winds, floods, blizzards, wildfires, and other weather-related events now cost hundreds of millions of dollars per day. Insurers need to make their policyholders whole, increasing costs.
The claims volume following an extreme weather event is usually much larger than normal. This presents a challenge that goes beyond claims processing, however: NICB research shows that up to 10% of disaster claims are fraudulent.
Fraud investigations following a disaster can be extremely sensitive. False positives can prevent policyholders from putting their lives back in order, and false negatives can raise insurers’ costs.
Going by the NICB rubric, a weather event that results in $1 billion worth of total claims may include $100 million worth of fraudulent claims. Insurers may not be able to shrug off these expenses – and if they can’t, they’ll be forced to pass these costs onto their policyholders in the form of increased premiums.
Weather events create large numbers of similar claims
Insurers will often find themselves swamped with similar-looking claims in the aftermath of an extreme weather event. Let’s look at roof damage – commonly associated with hurricanes, thunderstorms, tornadoes, etc. – as one example.
Per the NICB, if 20,000 policyholders file claims for damaged roofs after a storm, up to 2,000 of these claims may contain an element of fraud. The problem is that each claim involves the same kind of damage. If all the claims look the same at first glance, there’s no easy way for investigators to manually sort the suspicious claims from the legitimate claims.
Manually investigating all the claims is not an option. If the insurer investigates all 20,000 claims meanually, some of the policyholders will understandably resent the delay, and may churn. Meanwhile, simpler automated systems will end up paying a lot of fraudulent claims, again because there are few clear statistical markers that separate false and legitimate claims in the wake of a weather disaster.
Using AI to expedite legitimate disaster claims and flag suspicious ones
Enter AI fraud detection, which is more accurate, sensitive, and predictive than other solutions.
AI solutions can connect and utilize a large number of data sources related to each claim in real time. This makes it possible for AI to flag suspicious claims using elements that would be difficult for investigators to examine manually.
For example, AI could look at the 20,000 claims above and automatically flag claims where:
- Claimants have been previously suspected of fraud
- Contractors have been previously suspected of fraud
- Entities in claims have other suspicious links
- Photographic metadata shows damage that took place before the storm occurred
- Image recognition shows damage belonging to a different building than the one listed on a policyholder’s claim
Some Shift Technology clients even use satellite imagery to corroborate claims data. Here, Shift Technology incorporates external data from commercial satellite imaging companies. If a policyholder claims damage to their roof or yard, Shift can correlate their written description with their address as it appears on a satellite image. Shift will also use historical satellite images to ensure that the policyholder isn’t claiming unrelated damage – i.e., they’re not trying to pass a poorly-maintained roof as a storm-damaged roof.
Each analysis uses a data store and a method that would take an investigator much longer to perform manually. It could potentially take hours for an investigator to confirm that a policyholder actually owns the damaged building shown in their claim. Meanwhile, AI solutions have access to property records, satellite data, image recognition analysis, and more – and can take less than a second to perform a thorough analysis.
In short, AI can take a large number of similar-looking claims and narrow them down into a reasonable workload. This will include a prioritized list of potentially fraudulent claims, contextual guidance for investigators, streamlined investigations, better decisions, and improved outcomes for the insurer and its policyholders alike.
Using AI to Mitigate Systemic Risks from Extreme Weather
Insurers need to be fair to their policyholders when disaster strikes – while also ensuring that they can maintain their loss ratio without increasing premiums. That’s where Shift Technology comes in. We offer AI-native solutions that let insurers mitigate fraud and avoid customer churn. Even as extreme weather worsens, insurers can retain and grow their customer base while offsetting potential losses—the foundation of a strategy for climate resilience.
For more information on how AI can help insurers as they adapt to a changing climate, contact Shift Technology today.
