Technology

Industry-Leading Technologies

With the rise of generative AI, the only solution is to fight AI with AI. At Jumio, we continually research, develop and embed cutting-edge technologies into our solutions to stay at the forefront of fraud detection, compliance and UX.

Fraud Detection Technology

Document Analytics

Every government-issued identity document (ID) — whether it’s a driver’s license, identity card or passport — has a specification that defines the layout of the document, the relative position of specified fields, labels and images, font types, color, size, spacing and patterns.

Every ID also contains a unique set of security features such as a ghost image, watermarks, holograms, microprint, chip and machine readable zones. These features can vary from state to state and country to country.

Jumio performs a variety of AI-driven fraud checks on these ID documents to ensure that the submitted ID conforms to these government templates and does not exhibit signs of fraudulent tampering, such as text and photo manipulation. Using informed AI, our advanced machine learning models can detect sophisticated attacks that are undetectable by the human eye.

For supplemental documentation, such as when banks require a utility bill for proof of address, Jumio scans the document and extracts its data. It can also perform fraud checks to ensure the document wasn’t manipulated or forged.

image of California ID with bubbles circling photo.

Learn more about document fraud checks:

When the user takes a photo of their ID, Jumio scans the machine-readable zone (MRZ) and human-readable zone (HRZ) and uses optical character recognition (OCR) to extract the data. We apply various sanitation and verification checks to optimize accuracy, compare the MRZ and HRZ data, and compare the checksums with the original checksums for data verification.

On ID documents that contain a ghost image, we compare this to the main face photo to identify potential anomalies that are indicative of fraud.

Detects where two different ID documents are being used during the ID verification process, which can occur when a fraudster presents the front of one ID document and the back of another.

Detects whether the face photo on the ID document has been digitally or physically manipulated.

Checks for the physical punch hole many issuing agencies use to mark a license or passport as invalid, such as when you renew your license/passport and they punch the old one to invalidate it.

If a user takes a picture of an ID presented on another screen or uploads a screenshot of an ID from a phone, this can be an indication of fraud.

Compares the ID document to the known and specified dimensions and layout of a document sample.

Fraudsters will often make multiple attempts to defraud an organization. Jumio detects when the ID, face on the ID, or key personal identifiable data points have already been seen in an identity transaction that was rejected as fraudulent.

For ePassport documents, i.e. those that contain an RFID microchip, the Jumio Mobile SDK can detect the presence of the chip to then unlock it and extract the data and ID photo stored in it.

We detect the presence of a human face portrait on the ID document.

Ensure specific data fields comply with the standards defined by the issuing agencies.

Detects whether an ID image has been manipulated by looking for visual indications as well as evidence within a file's metadata.

If the user presents a copy of an ID document instead of capturing a live image of the document, this can be an indication of fraud.

On ID documents that contain microprint, we can detect anomalies that are potentially indicative of document manipulation.

Detects the presence of perforations (typically perforated text) on specific ID documents such as passports.

Detects when a deepfake face image has been pasted onto the ID.

Biometric Analytics

During onboarding, our face-based biometric technology compares specific facial features from the selfie — such as the distance between the eyes, nose, and ears — against those of the ID photo, ensuring they’re the same person. It also performs checks such as age estimation (does the person appear to be their stated age?) and detects whether the person is wearing a mask or is sleeping.

Learn more about biometric checks:

Ensures the person in the selfie is the same as the person who owns the ID.

Checks whether we have seen this face in previous identity transactions. If the data in the current transaction doesn’t match the data from the previous transaction, this feature rejects the current transaction.

Detects if more than one face is presented in the selfie.

Estimates the age of a person in the selfie. This can be used for age estimation or as a fraud fighting signal.

Analyzes the person’s eyes to ensure they are awake.
two square boxes. Box on top shows a black and white photo of biometric face. Behind box shows green gradient.
Image of two phone screens. Phone screen on left shows woman's face with green scanning circle. Phone screen on right shows Welcome screen with a blue check mark.

After onboarding, when the user returns to your platform, our biometric authentication solution can prompt the user to take another selfie and then compares it to the original biometric template to ensure this is the same person who onboarded.

Liveness Checks

Jumio’s advanced liveness detection technologies can detect sophisticated attacks including face swaps, face morphs and other advanced spoofing attacks. We leverage state-of-the-art AI to defeat these attacks and confirm that the person and their ID are physically present.

For example, we can determine whether the individual submits a screenshot or paper copy of a picture (i.e., a printed photo) instead of an authentic selfie — commonly referred to as a presentation attack. We can also detect video injection attacks, which involve injecting a deepfake into the digital stream by using a virtual camera or hacking the vendor’s API or SDK.

Learn more about liveness checks:

Checks if the selfie was presented on a paper printout or on a digital screen, such as when a fraudster points the camera at another mobile phone or a TV.

Detects when the user has held up a warped photo or other simple digital manipulation instead of a selfie.

Detects whether someone is wearing a 3D printed mask to bypass the liveness check.

Detects whether the face image in the selfie has been morphed.

Detects whether the selfie is black and white instead of color.

Checks if the face is partially covered (e.g., scarfs, medical masks).

Detects when the person used a pre-recorded video instead of a selfie.

Detects whether a face image has been swapped in the selfie.

Detects AI-generated selfies.

Data Analytics

Jumio uses advanced optical character recognition (OCR) to extract data from the user’s ID so we can compare it to the machine-readable zone (MRZ), calculate the user’s age for age verification, and more. We then check the data against a wide array of industry-leading, third-party data sources to provide an unmatched level of assurance and fraud protection to businesses worldwide.

Data checks include:

  • Device reputation
  • Linkage between name, email, phone number, and address
  • Phone number reputation
  • Email reputation
  • Social Security number and name verification
  • Checks against politically exposed persons (PEPs), sanctions and adverse media watchlists
  • Verification against local government agencies

These database lookups take place in real-time from the extracted data or as an individual completes a standard onboarding form. With device checks, the information is detected from the user’s device before they’ve entered any information at all. Jumio then dynamically tailors the user’s onboarding experience based on their risk profile, introducing additional checks only as needed and minimizing friction for legitimate users.

image of man wearing glasses and white t-shirt. Man is looking at screen. Right side of image has a green gradient overlay.

Predictive Analytics

Jumio transformed identity verification many years ago by inventing selfie-based identification, and now Jumio is once again ushering in a new approach with Jumio 360° Fraud Analytics. This groundbreaking technology uses AI-driven predictive analytics to identify fraud patterns across our cross-enterprise network and accurately predict the likelihood of fraud whenever a user goes through our ID verification process.

This approach goes beyond assessing the user’s ID and selfie and now adds sophisticated behavioral analytics to determine how each identity transaction fits within our vast network. This makes it simple to spot much more complex connections and is especially effective at detecting fraud rings and preventing fraud. Rather than just looking back to see where fraud has already happened, this approach can stop fraud before it happens — automatically.

Jumio 360° Fraud Analytics is currently available as a preview release for select customers and will be fully available in 2024.

Image of circles with boxes on the inside.

Platform Technology

image of male figure waring glasses touching screen with biometric lines popping up.

The Jumio Platform provides a complete end-to-end solution for onboarding and ongoing monitoring of your users. You can orchestrate dynamic workflows that trigger identity verification, risk signals and AML screening services during the user journey based on the risk level of each user, minimizing friction for legitimate users. The real-time rules editor allows you to respond immediately to changing fraud trends. And the powerful analytics dashboard lets you easily spot trends in transactions, usage and success metrics and run detailed reports to help you maximize efficiency.

During the user journey, the AI-driven results from the various fraud checks feed into a single, unified risk score. The score calculation is configurable for your business and can be used in your workflows to trigger additional checks as needed.

One vendor. One API layer. One extremely powerful, highly secure and configurable platform. Jumio provides all of the data, business rules, orchestration, analytics and supporting tools you need to manage fraud across the user journey.

Integration Channels

Following are the integration channels you can use to implement the Jumio user journey. To try out the user journey, download the Jumio Showcase app for iOS or Android.

Web Client icon
Web Client
Our Web Client makes it easy to embed the user journey into your existing onboarding workflow for a seamless experience. It provides a default user journey as a web app that runs inside a web browser on all types of devices. The Web Client provides clear guidance to your end users and manages the interactions with their camera and other device features. You can make basic customizations such as changing the colors of the screens and buttons. The Web Client is hosted by Jumio, so it requires no maintenance.
Web SDK icon
Web SDK
If you want to fully customize the user journey and host it yourself, the Web SDK is a great option. You can stand up an instance of the Web Client with just a few lines of code and then have complete control over its look and feel, integrate it with your analytics framework, and modify the user journey to meet your needs. It’s framework agnostic, so you can use your preferred IDE. You also have full control over updates and other maintenance.
apple and android icon
Mobile SDKs for Android and iOS
Jumio’s mobile SDKs provide everything you need for a fully customized, mobile-native experience for Android and iOS. We include plugins for React Native, Flutter and Apache Cordova.
API icon
API
The REST APIs are used for uploading credentials and finalizing the workflow. Your application can call the APIs directly when you need to control other aspects of the customer journey.
multi-platform icon
Omnichannel Support
Jumio’s omnichannel approach makes it easy for users to complete the identity verification journey on their computer, on their mobile device, or a combination of the two. This is especially important in geographies where end users often have desktop computers with older-model webcams that offer poor image resolution compared to modern smartphones. These users can start the user journey on their desktop and then switch to their mobile device simply by opening an SMS message or scanning a QR code. After taking the photos on their mobile device, they can return to their desktop computer to complete the journey.

To learn more about the integration channels, see the Jumio Integration Guide.

DISCOVER MORE

image of woman with biometrics lines on face.
METRICS THAT MATTER
How to Evaluate Identity Verification Technology
image of woman with her hair up wearing earrings. Woman is pointing her finger to a screen.
Read the latest updates
from our Engineering and Data Science teams.

Get Started

Let a Jumio expert show you how easy it can be to integrate our automated identity proofing solutions into your existing processes.
image of man with facial hair smiling wearing a suit.
Before you go, may we suggest...
Metrics That Matter
How to Evaluate Identity Verification Technology