Unmasking Cybercrime Strengthening Digital Identity Verification against Deepfakes 2026
Page 4 of 23 · WEF_Unmasking_Cybercrime_Strengthening_Digital_Identity_Verification_against_Deepfakes_2026.pdf
FIGURE 1 Anatomy of a deepfake attackBackground and motivation
Face swapping is an AI-enabled form
of impersonation and identity theft.01
The term deepfake refers to synthetic audio or visual media
that convincingly imitates real people, created through
machine learning methods (typically deep neural networks).2
This technology enables the creation of highly realistic but
fabricated images, videos and audio recordings that are often
hard to distinguish from authentic content.3 Originally developed
for entertainment and creative applications, deepfake
technology has increasingly been exploited for malicious
purposes, including fraud, disinformation and impersonation.4One prominent technique within the domain of deepfakes is
face swapping.5 Face swapping involves replacing the facial
identity of an individual in an image or video with that of
another person, while preserving expressions, movements
and contextual realism.6 When misused, face swapping
can serve as a sophisticated mechanism for identity theft,
enabling attackers to engage in unauthorized activities
under the guise of a victim’s likeness.7What are deepfakes and face swapping?
Stolen or AI-
generated identity
documents
are prepared.High-quality
face-swapping tools
are used to generate
face-swapped media
that matches submit-
ted documents.Camera injection
tools are used to
feed synthetic
video into live
biometric checks.Cybercriminals
bypass automatic
identity verification.
Unmasking Cybercrime
4
Ask AI what this page says about a topic: