Liveness detection
Also known as: Liveness check · Presentation attack detection · PAD · Anti-spoofing
Verifying that the person in front of the camera is present and alive at the moment of capture — not a photo, a screen replay, a mask or a synthetic video.
Legal basis
ISO/IEC 30107-3:2023 (presentation attack detection — testing and reporting)
Liveness detection is the layer that stops someone from impersonating another person by holding that person's image up to the camera. It can be active, asking for a gesture — blink, turn your head — or passive, analyzing texture, depth and reflections in a single capture without the user noticing. Passive reduces drop-off; active is easier for users to understand.
The reference standard is ISO/IEC 30107-3, revised in 2023, which sets how presentation attack detection is tested and reported and classifies known attack types. Vendors cite conformance testing against it; the standard itself covers attacks presented to the capture device, not system-level security.
That boundary matters because the threat has moved. FinCEN warned in 2024 that criminals use generative AI to defeat identity verification, and injection attacks feed synthetic video straight into the stream without any camera. Liveness confirms presence; it does not answer whether that real person is the one who should be signing — that requires cross-checking the identity against other sources.
Frequently asked questions
What is liveness detection?
It is a check that the face presented to a camera belongs to a live person who is physically present, rather than a photo, screen, mask or video, typically run during remote identity verification.
What is ISO 30107-3?
An international standard for testing and reporting the performance of biometric presentation attack detection. Liveness vendors are commonly evaluated against it by independent labs.