Fraudsters using AI to manipulate images for false claims

Insurance fraud costs the property and casualty insurance industry about $45 billion annually, according to the White Collar Crime Research Task Force at Colorado State University Global. These losses include homeowners, auto and commercial insurance, and American families foot some of the bill, paying up to $700 more per year to cover fraudulent claims and inaccurate applications. Now, fraudsters are using AI to manipulate images for fake claims. AI-related insurance fraud is up by 300% in all areas of insurance, according to an investigation by Allianz UK.

“Just as we use technology to help us detect fraud, dishonest people are increasingly using technology to try to get us to pay out for something that may not even have happened,” said Scott Clayton, head of claims fraud at Zurich UK. “We, as insurers, have to keep up with technology and be alive to some of the new and emerging ways that people will try to make a fraudulent claims [sic].”

“Shallowfakes”

Photo editing apps with generative AI features have made it easy to doctor images and videos, creating what experts call “shallowfakes” or “cheapfakes.” These manipulated photos may appear legitimate at first glance and show how desperate scammers adapt their methods using new technology.

One fraudster in the UK filed a claim for an accident that never happened, using an image of a cracked bumper on his work van and a fake repair bill totaling more than $1,900, or €1,000. His insurer, Liverpool Victoria (owned by Allianz), had concerns about the case and sent the images to its fraud team for investigation. They found an identical image on the man’s social media account, without vehicle damage.

An investigator at Aviva foiled another fraud attempt involving a $21,000 watch (£20,000) by discovering an identical photo online that showed the exact time down to the second. Aviva found a reported 9,250 fraud cases in 2022, helping the insurance giant avoid more than $129 million (£120 million) in false claims.

Insurers fight back

This growing activity in AI-related insurance fraud highlights the new ways fraudsters are targeting insurance companies and illustrates the need for insurers to step up their fraud detection. Insurance fraud evolves, so must robust controls to spot these emerging trends, says Matt Crabtree, head of financial crime intelligence and investigation strategy at Allianz. He shares that insurers need to “root out” those trying to exploit the system to protect honest policyholders and keep premiums at a minimum.

“Protecting insurance customers against the impact of fraud remains an industry priority. Insurers are alive to the various tactics fraudsters deploy and use sophisticated software to detect false or edited images,” said Mark Allen, chief fraud and financial crime officer at ABI. “The use of technology is a key part of the insurance sector’s counter fraud weaponry and there will be no let-up in the industry’s prevention and detection of fraudulent claims, and enforcement against insurance fraudsters.”

Related:

• Lights, camera, auto fraud: Scammers caught on video staging accident
• Who commits insurance fraud and why
• Cities that saw the most insurance fraud in 2023