The World’s First Standalone Document Liveness Detection Product
IDLive Doc is the first standalone product for document liveness detection, an essential countermeasure to protecting KYC and digital onboarding from document-based presentation attacks.
IDLive Doc builds on ID R&D’s deep expertise in single-image passive facial liveness, which detects the use of photos, screen replays, and masks during facial matching.
Recent ID R&D research shows that over 90% of all document-based attacks on digital onboarding are presentation attacks.
Bad actors present digital images of identification documents instead of “live” IDs in their possession to spoof document authentication functions and commit fraud. Preventing this type of attack is essential to a rigorous document verification process. Document liveness detection uses AI and computer vision to distinguish between a present document and a digital image shown on a mobile device or computer display to prevent this type of presentation attack.
IDLive Doc from ID R&D detects digitally-displayed document spoofs. The product builds on ID R&D’s deep expertise in single-image passive facial liveness, which detects the use of photos, screen replays, and masks during facial matching. Production deployments of IDLive Doc have revealed that document spoofing with digitally-displayed images is on the rise and more common than realized.
ID R&D’s IDLive Doc saved RealAML years of in-house development, helped acquire new customers, and immediately demonstrated its ability to stop fraud attacks. IDLive Doc integrated seamlessly into our existing KYC process without disrupting other core components.
Banks and other institutions increasingly allow customers to set up accounts remotely using a mobile onboarding app in a process called digital onboarding. Also known as electronic Know Your Customer or eKYC, the process requires digital identity verification and typically includes the presentation of a identity document, such as a passport, national ID, or driver’s license, as well as capture of a “selfie”. It’s a great convenience to customers and generates more revenue for businesses, but creates an opportunity for fraud. Fraudsters can easily source digital images of identity documents from illicit sources, and then manipulate them digitally. This is why it’s so critical to determine…is the document physically present? Or, is a fraudster presenting an digital image of a document displayed on their mobile phone or computer monitor?
Identity document fraud using a digitally-displayed image is called a document screen replay attack. Screen replay attacks are accelerating, yet many onboarding systems cannot reliably detect them. Often the spoofed document is not discovered until much later in the customer lifecycle when other red flags trigger an audit. IDLive Doc detects a screen replay at the time of the attack, creates more risks and costs for bad actors, and prevents fraud before it causes damage. Importantly, it does so without adding friction in the user experience for legitimate customers, and without tipping off fraudsters that the countermeasure is in place and how it might be defeated.
IDLive Doc is easily added to any existing remote onboarding system to stop screen replay attacks of identity documents. Once the IDLive Doc Docker image is installed, all that is needed is one API call with a payload of the same image already captured for identity document verification. IDLive Doc provides the probability of document liveness and you choose how to respond if a screen replay attack is detected. You can be up and running in hours. IDLive Doc works universally across all types of identity documents from around the globe, including driver’s licenses, passports, and national IDs.
Minimize the number of attack vectors in the digital onboarding process with a layered approach to document verification that involves document authentication, document liveness, facial recognition, and ID R&D’s IDLive® Face passive facial liveness detection.
The document liveness check is performed in the background in milliseconds, with no incremental friction for the customer or warning for fraudsters.
