Unmasking Cybercrime Strengthening Digital Identity Verification against Deepfakes 2026
Page 7 of 23 · WEF_Unmasking_Cybercrime_Strengthening_Digital_Identity_Verification_against_Deepfakes_2026.pdf
A multi-stage flow is commonly followed when face swapping
is used to bypass KYC:
1. Preparation and reconnaissance: Targeting criteria
are defined on either bulk account creation or
targeted impersonation.
2. Document misuse or forging: A document is either
stolen, purchased, forged or AI-generated to match
the target identity.
3. Face swapping and synthetic media : A face swap model or
pipeline is trained or configured to create synthetic images
or videos that match identity documents. These tools may
provide both pre-generated and real-time synthetic media.144. Camera feed substitution: Camera injection tools are
used to replace a device’s default live camera feed with
AI synthetic media. When the KYC system captures the
“live” biometric, it receives the forged media instead of the
actual user, allowing the attacker to pass biometric checks.
5. Device manipulation: Emulators, virtual environments
or other device manipulation techniques are rotated
or re-provisioned to avoid linking multiple account
registrations to a single physical resource.
6. Network layering: Residential proxies, mobile proxies or
VPNs (virtual private networks) are used to diversify IP
(internet protocol) addresses and geolocation signals,
and to defeat simple IP-based correlation. Detailed multi-stage attack flow
Unmasking Cybercrime
7
Ask AI what this page says about a topic: