Injection attack
Feeding a fake image or video directly into a verification flow, bypassing the camera.
In an injection attack, the fraudster doesn't hold anything up to the camera. Instead, they insert pre-recorded, edited or AI-generated footage into the app or browser, for example through a virtual camera driver, an emulator or a modified device.
Why they are dangerous
Because the footage never passes through a real camera, checks that only look for screens, paper edges or masks in the image can miss it. Combined with deepfakes, injection attacks are one of the main threats to selfie-based verification.
How they are detected
- Checking that the capture comes from a genuine device camera.
- Detecting emulators, virtual cameras and tampered apps.
- Analysing the video itself for signs of synthesis.
- Linking repeated attempts across sessions.
Contact
Talk to us
Tell us what you’re trying to verify. We’ll show you how secuveri can fit your workflow.
