Certified against deepfakes,
injection attacks, and silicone masks.
As AI-generated synthetic media becomes indistinguishable to the human eye, liveness detection is no longer optional, it is the primary defence line for any identity verification flow. WeVerify's biometric matcher is certified under ISO 30107-3 and produces a 0% imposter accept rate in live production deployments.
What we detect
Every category of presentation and injection attack.
AI deepfake video
Real-time detection of AI-generated face-swap video presented during the selfie step. Detects even high-quality GAN and diffusion model outputs.
Digital injection attacks
Detects pre-recorded or synthetic video injected directly into the camera API, the most common professional fraud attack vector today.
Printed photo attacks
Detects static printed or screen-displayed photos held in front of the camera during the liveness check.
Silicone mask attacks
Detects 3D silicone masks and face prosthetics, a high-cost but increasingly accessible attack vector for high-value identity fraud.
Certification
ISO 30107-3 Level 2 compliant. 0% IAPAR in production.
How it works
Passive liveness, no motion challenges.
WeVerify uses passive liveness detection. There are no instructions to blink, smile, or turn your head. A single selfie is analysed for liveness indicators invisible to the human eye, texture inconsistencies, reflection patterns, and temporal anomalies in video. This reduces drop-off and removes the hint to fraudsters about what to defeat.
The NFC advantage
Liveness alone isn't enough. A fraudster with a deepfake video and a stolen passport photo can defeat liveness-only systems. WeVerify combines liveness with NFC chip reading, the biometric extracted from the chip is matched against the selfie. The fraudster would need both the real passport in hand AND the synthetic media to match it. The combination eliminates this attack vector.
Deepfake-certified identity verification.
Book a demo and we will run a live liveness check, explain the detection layers, and show you the evidence output.