Glossary term
Injection Attack
Definition
An injection attack bypasses the physical camera or microphone and feeds pre-made or real-time synthetic media directly into the verification data stream, using virtual cameras, tampered apps, or intercepted API traffic. Unlike a presentation attack, nothing is shown to a sensor — the fake enters the pipeline as data.
Why it matters for financial institutions
Injection attacks defeat controls that assume a trusted capture device, and industry analyses now report them growing several times faster than presentation attacks against video KYC. For fintechs and crypto platforms with fully automated onboarding, an injected deepfake stream can pass both document matching and liveness without a camera ever being involved.
How FalsiFind addresses it
Because FalsiFind analyzes the media content itself rather than trusting the capture path, injected deepfakes are scored with the same generation-artifact analysis as any other media. Suspicious streams are flagged regardless of whether they arrived through a real camera or a virtual one.