Verisk Launches AI-Powered Fraud Detection Analytics to Identify Fraudulent Images

The new forensic tool lets insurers quickly identify when and where photos of damages were taken, and if they’ve been used in previous claims, copied from websites or altered digitally.

(Example of articles of clothing similar to the ones involved in a claims fraud case. Image source: Verisk.)

Verisk (Jersey City, N.J.), a global data analytics and technology provider, has expanded its anti-fraud claims solution with image fraud detection analytics, powered by artificial intelligence, which the vendor says that will enable insurers to identify potential fraud in claims photos quickly and precisely.

Drawing on millions of claims images in Verisk’s databases, including ClaimSearch loss history database, the image forensics solution analyzes every image submitted as part of a claim to detect indicators of fraud, the vendor reports. Verisk says that, without any human intervention, the solution automatically flags photos that have been altered, manipulated or downloaded from the internet. Using ClaimSearch, Verisk’s solution also identifies images that have been submitted to other insurers.

Verisk provides an example of the efficacy of the new forensic capabilities with the case of a claim filed by a policyholder for the theft of jeans, shows and a purse with images he said were taken a few months earlier. Using the images’ metadata, Verisk’s ClaimSarch determined that the photos were actually taken the day after the alleged theft, according to the vendor. “Fraud was prevented and the ‘thieves’ were caught, all thanks to image forensics,” a Verisk statement says.

Digital Transformation of Insurance Fraud

Reusing or altering photos is just one example of the growing trend of insurance fraud, which costs the United States approximately $308.6 billion annually, according to the Coalition Against Insurance Fraud.

Shane Riedman, General Manager, Anti-Fraud Solutions, Verisk.

A Verisk study of a sample group of 768,000 claim images found 1,967 duplicates, including one photo used in 44 different claims.  These duplicate claim images could be linked to 1,475 separate claims for which approximately $5.3 million in indemnity payments were made.

“The ubiquity of mobile devices, photo editing software and claims filing applications is making it easier than ever to submit fraudulent images of damages to insurers,” comments Shane Riedman, VP and general manager of anti-fraud analytics, Verisk Claims Solutions. “The image analytics and alerts we’re adding to ClaimSearch gives insurers the ability to quickly, automatically identify and investigate suspect claims, while simultaneously accelerating meritorious claim resolution for the vast majority of policyholders.”

Insurers who contribute claims data to ClaimSearch—currently more than 95 percent of the industry—can submit claims photos and access certain image forensics checks with their current ClaimSearch subscription. To date 315 insurers have signed up to contribute images, and nearly half of the top 20 property/casualty carriers are accessing the solution today, Verisk reports.

Verisk Acquires Morning Data to Support Digital Transformation of Brokers and MGAs

Anthony R. O’Donnell // Anthony O'Donnell is Executive Editor of Insurance Innovation Reporter. For nearly two decades, he has been an observer and commentator on the use of information technology in the insurance industry, following industry trends and writing about the use of IT across all sectors of the insurance industry. He can be reached at or (503) 936-2803.

Leave a Comment