Visualize: Insights that power innovation

Visualize: Insights that power innovation

Estimating claims remotely? Uncover potential fraud in digital loss images

By Joshua Louwagie  |  April 15, 2020

The COVID-19 outbreak has had a tremendous impact on every aspect of life in the United States. Like most companies, the insurance industry has had to drastically alter the way it does business. Claims adjusters are now working at home, relying on customer-submitted images and videos to inspect and appraise damages for the foreseeable future. And even after the pandemic abates, the disruption the virus has caused will likely result in permanent shifts in the way carriers investigate claims.

While virtual loss inspections have their advantages—such as improving operational efficiency and customer experience—they also present challenges with fraud detection that the entire industry needs to address.

Reevaluating the risk of virtual claim inspections

Insurers have made great strides in recent years in adopting photo-based estimates as the industry has become more digital. Consider this: Photo estimates for auto claims increased tenfold from 2016 to 2017. Consumers have also embraced the technology, as J.D. Power reports 42 percent of claimants submit vehicle loss images or videos to insurers.

Carriers have primarily used photo estimates for simple, low-value claims—such as auto glass losses or minor property damage—that they automatically pay if it’s under a certain threshold. But with stay-at-home orders in effect across much of the country, insurers are forced to expand to claims with larger loss values. And that can be a game changer.

The fact is, insurers can’t afford to auto-pay large losses, and the industry doesn’t have the necessary checks in place to detect fraud associated with digital images.

Detecting digital image fraud is difficult

Customer-submitted and third-party loss photos already potentially expose carriers to fraud, such as someone submitting the same loss image to multiple insurers. But the scale at which adjusters are conducting photo-based estimates during the crisis—combined with the rising threat of fraudsters trying to profit from the pandemic—results in the increasingly greater risk of paying costly, illegitimate claims.

Currently, the process for detecting preexisting damage or fraud in digital images is generally manual, labor-intensive, and inefficient. Adjusters must contact other carriers to request photos of prior losses and wait to receive the images, which delays the claims process. With increasing workloads and dwindling staff, claims teams don’t realistically have time to go through this procedure for every photo-based estimate.

Identifying fake images—such as a photo downloaded from the Internet—is even more challenging. Most adjusters aren’t trained or equipped to detect suspicious images, and available online tools don’t provide sufficient analysis to uncover fraud.

Industrywide image data critical to anti-fraud efforts

Detecting preexisting damage and digital image fraud at scale requires sophisticated technology and extensive data from across the industry. It also takes an industrywide effort because no carrier can do it alone.

While an insurer may be able to identify preexisting damage from prior losses in its book of claims, there’s no efficient way to detect if loss images have been previously submitted to other carriers. That requires a combination of industrywide loss history records and prior-loss images that can be analyzed and matched.

New investigative tools automate image fraud detection

Verisk is helping the industry meet virtual claims-handling challenges with a new, digital media database. This contributory database is an enhancement to ISO ClaimSearch®, the industry’s largest loss history database, that allows users to easily access prior-loss images on match reports for free—and without any IT lift.

Carriers can use already-established APIs to upload customer or claimant submitted images to a cloud-based repository when they report claims to ISO ClaimSearch. The platform will then automatically indicate whether the image data corresponds with prior claim information or if a duplicate photo already exists in the database.

This enhancement also includes cutting-edge image forensics technology that searches the Internet for any photos that match files submitted with a claim and analyzes metadata to determine if the image data (date, time, and location) corresponds with the claim information.

Imagine how these tools can help insurers improve efficiency and reduce leakage: Claim handlers can quickly and easily check for preexisting damage within their workflow and get insights needed to either expedite legitimate losses or investigate suspicious claims.

Mitigate new risks effectively

The industry is undergoing a sea change accelerated by the COVID-19 pandemic. As photo-based estimates become the new norm, it’s critical to have fast and efficient fraud checks in place to mitigate risks. The digital media database and image forensics tools enable insurers to do just that.

By adding digital image fraud detection capabilities to its robust data and sophisticated claims-matching technology, ISO ClaimSearch is equipping the industry to get ahead in the fight against fraud.


Joshua Louwagie is a senior product innovation manager at ISO Claims Solutions, a Verisk business. You can contact him at joshua.louwagie@verisk.com.