Incognia Launches Deepfake Detection Tool

Incognia has released a new fraud prevention tool to strengthen liveness checks during onboarding. In that regard, the location-based Liveness Spoofing Detection module is designed to be used alongside a face-based onboarding solution that asks users to submit a selfie and a picture of an ID for identity verification.

As it stands, most selfie onboarding systems claim to use liveness detection, which is to say that they analyze the selfie photo (or the selfie video) to make sure that a real person is present. The problem is that many of those liveness detection systems simply aren’t up to the task. Fraudsters have been able to create convincing deepfakes with facial images pulled from the web, and then use those deepfakes to get around liveness evaluations.

In some cases, the deepfake is presented directly to the camera (a presentation attack), while in others, the fraudster carries out an injection attack to bypass the camera entirely and submit a pre-existing video instead of a live one. Incognia believes that its new solution can address both threats.  To do so, the Liveness Spoofing Detection module analyzes the device itself to watch for signs of tampering, or the use of emulators. The solution will send up a red flag if a device has been rooted or jailbroken, or if other risk indicators are detected.

The Spoofing Detection module can also be paired with other Incognia modules for a higher level of security. The company’s flagship Location Spoofing Detection module looks at a device’s location data and warns businesses if someone tries to log in from a place that deviates from their normal path, while the Global Mobile Address Validation module cross references someone’s movement patterns with the address submitted during onboarding to determine whether or not the individual actually lives at that address.

According to Incognia, the new solution can be used in a number of different sectors, ranging from financial services to social networks. The module provides a degree of behavioral analytics, and can check device data against watchlists of devices that have previously been associated with fraudulent activity. Incognia itself brought in $15.5 million in Series A funding earlier this month.

–

June 14, 2022 – by Eric Weiss