Deepfake detection software helping insurers fight fraud

With the ease of creating convincing deepfake images and even videos threatening to increase insurance fraud, a Utah-based software company is equipping insurance companies with another layer of anti-fraud protection through automatic deepfake detection.

“With this whole proliferation of generative AI, it’s very easy for people to create fake photos or videos. That’s one of the aspects that our software does very well — it identifies and flags photos and videos that were created with gen AI,” Nicos Vekiarides, co-founder and CEO, Attestiv Inc. said.

Those anti-fraud capabilities translate into a huge ROI for insurers, considering fraud already costs insurers billions every year and the threat of deepfakes is increasing as gen AI becomes more easily accessible.

“If you think about Photoshop, that’s kind of the realm of a graphic artist, maybe a hacker. But with the types of things you can do with generative AI today, you can go to Dolly or Stable Diffusion or Midjourney and you can create pictures, you can take existing photos and make some changes to them, exaggerate damages. All of that has really raised the threat of fraud, particularly in processes that are automated,” he explained.

The increasing threat makes deepfake detection software such as that offered by Attestiv even more valuable for insurtechs, which are in the business of automating more parts of the insurance process.

“When you’re automating things, this is where it becomes very easy to lose sight of the fact that the photos can be fake, videos can be fake, all of the data that you’re processing can be fake. It becomes almost imperative that now you have to really take some precautions against this threat,” Vekiarides said.

Using AI to detect AI

Attestiv’s solution uses AI technology to detect photos, videos and documents that have either been created entirely using AI or manipulated in some way. The company aims for data precision in the high-90s, and frequently updates its data models.

The online platform ranks photos, videos and documents on a score of 0 to 100. Anything scored over 50 is automatically flagged as suspicious for the insurer to take further action.

It’s also offered as an API, where companies can integrate it into apps that even clients can use. Vekiarides believes this is where the future of such technology is headed.

“How do you make it really easy for customers to consume this technology? Well, integrate it into applications that they already use… Take advantage of self-service so that the policyholder can take photos in the fender bender or so forth. That really enables insurance companies to get this instant fraud identification or instant identification of suspicious items in a claim without necessarily making any big changes to their process,” he said.

Companies already taking advantage

According to Vekiarides, an increasing number of insurtechs have expressed interest in Attestiv, and the company is preparing for “a pretty large launch with a claims provider.” It already partners with companies like Repwest, Insurity and PCMS.

Repwest provides the insurance for well-known moving company U-Haul.

“With us, they’ve launched an online process where the renters or claimants can submit their photos online. Once they get submitted online, they first go through the Attestiv process and if we find something suspicious within a claim, then it gets flagged for further attention. If we don’t, then it can go through a little bit of a faster claims processing,” Vekiarides explained.

Insurity and PCMS, meanwhile, have integrated Attestiv’s technology directly into their existing claims systems through APIs.

“Insurance carriers who use these systems don’t need to change their processes anymore. They can simply run their processes the same way they did before, except they will get an Attestiv report with every claim that tells them whether there is anything that we found that’s suspicious in the claim,” Vekiarides added.

‘Enhancing’ the insurance ecosystem

Michael Desrochers, VP, strategy & growth, PCMS, said his company sees great value in the “unified technology solution” offered by Attestiv.

“Where I see this technology being very beneficial is to the actual company or insurance entity that is going to be writing the risk, handling the risk. So, if you are a carrier, an insurance carrier, I think it’s great from the underwriting perspective to take more preventive measures,” Desrochers said.

For instance, he noted specific instances where it could be beneficial for P&C purposes, such as to insure an expensive jewelry item or verify damage claims after a hurricane. However, he suggested it could even benefit a life insurance agent, as a document-tampering detection ability could be used to validate documents such as wills.

“Having the ability to leverage AI-powered technologies to prevent fraud or to detect fraud is really a value to us as a solution provider to these carriers. Having someone like Attestiv integrated in our solution really enhances our ecosystem, and it is an augmenting capability piece,” he said.

Attestiv is a private software company founded in 2018 and based out of Lehi, Utah. Its cloud-based platform aims to ensure authenticity of digital assets.

Property & Casualty Management Systems, Inc. is an American insurance company founded in 1999. Based in Dallas, Texas, it offers P&C insurance software solutions to small and medium-sized carriers.

© Entire contents copyright 2024 by InsuranceNewsNet.com Inc. All rights reserved. No part of this article may be reprinted without the expressed written consent from InsuranceNewsNet.com.

Rayne Morgan is a journalist, copywriter, and editor with over 10 years' combined experience in digital content and print media. You can reach her at [email protected].