Jumio Leverages AI to Fight Fraud with First-to-Market Video Selfie Authentication Solution

Additional Info

Company size (employees)1,000 to 4,999
Type of solutionSoftware


In February 2019, Jumio launched Jumio Authentication, an AI-powered video-selfie authentication solution that was first in the market to leverage face-based biometrics for initial identity proofing and ongoing user authentication. Jumio Authentication allows modern enterprises to prove user identity in online transactions by unlocking their users’ digital identities in seconds. This secure and rapid authentication method is ideal for account logins and high-risk scenarios (e.g., logging in from a foreign IP address or authorizing high-risk transactions such as wire transfers and online purchases).

To help stay ahead of fraudsters, Jumio integrated the world’s first and only iBeta Level 1 and Level 2 certified anti-spoofing technology into its online identity verification suite to capture and process the user’s face images taken from any 2D smartphone camera or webcam. The face frames taken with Jumio Authentication leverage AI to create a 3D face map which contains over 100 times more liveness data than a 2D photo. This type of liveness detection ensures the user is physically present and prevents masks, dolls, photos, or videos, including deepfakes, from fooling the system. These liveness detection algorithms must observe so many concurrent human traits that no spoof can recreate them all at once.
Jumio also surpassed the 200 million identities verified milestone in 2019 and is verifying 300,000 identities daily. This has equipped Jumio with a significant advantage in training its AI algorithms to more accurately identify and detect fraud patterns. Using the company’s massive database of historical identity verifications, AI scientists at Jumio’s AI Labs are developing deep learning models to improve identity verification accuracy and speed. In 2019, Jumio’s AI models have reduced the average customer transaction time by more than 50% and sped up the authentication process; Jumio can now deliver a definitive match/no match determination in 2-3 seconds.

How we are different

Jumio can automate over 90% of its face match solution based on the data from authentication and 3D liveness products, allowing them to quickly and reliably make faster identity verification decisions to support more than 3,500 global ID types and subtypes.
Jumio’s scale and deep experience in ID and identity verification have amassed very large datasets that are being leveraged to develop highly predictive AI models. Jumio uses AI to extract key data from ID documents. Jumio’s scale means they can feed their algorithms with lots of data to not only improve their ability to recognize specific ID documents, but also to know how to extract the data and make sense of it.
AI also helps Jumio spot characteristics of fraudulent IDs, such as understanding the unique fonts, pictures, and security features of a specific country-issued ID. If an ID document has been manipulated or changed and does not conform to the pattern, Jumio’s AI algorithms flag it for closer review. Jumio also leverages AI to identify patterns and probability of fraud based on a combination of high risk variables. These insights have allowed us to identify that 1% of the ID documents (with a unique risk profile) account for about 15% of known fraud. We use this scoring to alert our verification experts to pay special attention to high risk IDs.