A separate ARTEX-associated workstream targeted verification at two cryptocurrency exchanges using generated video and forged identification images. Its agent records describe reverse-engineering a liveness vendor's face-motion challenge and testing real-time animation. The best logged synthetic face tracked the expected location within seven pixels; an…
Inspect the ClaimsCase · DiggingBeagle record
ARTEX agents reproduced face-movement prompts, but recorded KYC checks rejected every synthetic attempt
A separate ARTEX-associated workstream targeted verification at two cryptocurrency exchanges using generated video and forged identification images. Its agent records describe reverse-engineering a liveness vendor's face-motion challenge and testing real-time animation.
Public disclosure: Nov 13, 2026
Offensive agents and facial animation tooling can reach or emulate exchange onboarding and identity verification steps.
At a glance
Following a visible client-side movement challenge is distinct from passing the server-side identity decision and authorizing a withdrawal.
Read the implicationsThe report draws on an attacker-side decoy interaction. Exchange audit logs, real accounts and any successful cashout were not independently confirmed; the KYC study must not be merged into the separate account-key harvesting operation.
Limits and uncertaintyFull account
A separate ARTEX-associated workstream targeted verification at two cryptocurrency exchanges using generated video and forged identification images. Its agent records describe reverse-engineering a liveness vendor's face-motion challenge and testing real-time animation.
The best logged synthetic face tracked the expected location within seven pixels; an optimized consumer-GPU pipeline reached approximately 11.5 frames per second. The attackers also manipulated JPEG camera metadata and compression evidence, seeking consistency with photographs rather than screen captures.
The server-side liveness verdict was REJECTED for every attempt shown in the recovered sessions. This is evidence of a coordinated attacker effort and a measured failure of those attempts, not evidence that an exchange authorized a forged identity or released funds.
Reported figures and denominator. The strongest movement attempt was within seven pixels; measured generation was roughly 11.5 frames per second; server-side accepted sessions in the recovered record were zero.
The report draws on an attacker-side decoy interaction. Exchange audit logs, real accounts and any successful cashout were not independently confirmed; the KYC study must not be merged into the separate account-key harvesting operation.
Mechanism and trust boundary
- 01
Starting condition
Offensive agents and facial animation tooling can reach or emulate exchange onboarding and identity verification steps.
Boundary: precondition and attacker capability
- 02
Trust boundary crossed or evaluated
The best logged synthetic face tracked the expected location within seven pixels; an optimized consumer-GPU pipeline reached approximately 11.5 frames per second. The attackers also manipulated JPEG camera metadata and compression evidence, seeking consistency with photographs rather than screen captures.
Boundary: mechanism
- 03
Observed consequence
Following a visible client-side movement challenge is distinct from passing the server-side identity decision and authorizing a withdrawal.
Boundary: observed effect and limitations
Claims & evidence
CLM-NOV15-ARTEXKYC-MECHA separate ARTEX-associated workstream targeted verification at two cryptocurrency exchanges using generated video and forged identification images. Its agent records describe reverse-engineering a liveness vendor's face-motion challenge and testing real-time animation. The best logged synthetic face tracked the expected location within seven pixels; an optimized consumer-GPU pipeline reached approximately 11.5 frames per second. The attackers also manipulated JPEG camera metadata and compression evidence, seeking consistency with photographs rather than screen captures.supported
Basis: reported finding
- supportsAmazon Threat Intelligence: ARTEX agent cybercrime campaignsprimary disclosure
Amazon Threat Intelligence, Campaign 2: Defeating identity verification with deepfakes
CLM-NOV15-ARTEXKYC-SCOPEThe server-side liveness verdict was REJECTED for every attempt shown in the recovered sessions. This is evidence of a coordinated attacker effort and a measured failure of those attempts, not evidence that an exchange authorized a forged identity or released funds. The report draws on an attacker-side decoy interaction. Exchange audit logs, real accounts and any successful cashout were not independently confirmed; the KYC study must not be merged into the separate account-key harvesting operation.supported
Basis: reported finding
- supportsAmazon Threat Intelligence: ARTEX agent cybercrime campaignsprimary disclosure
Primary disclosure: results, scope and limitations
Implications
Following a visible client-side movement challenge is distinct from passing the server-side identity decision and authorizing a withdrawal.
Controls and mitigations
- Bind KYC verdicts and subsequent withdrawal permissions to server-side challenge receipts, replay detection and session identity.
- Treat visual motion scores, synthetic EXIF and generator throughput as attacker capability measurements, not as accepted-user or financial-loss counts.
Unknowns and contradictions
- The report draws on an attacker-side decoy interaction. Exchange audit logs, real accounts and any successful cashout were not independently confirmed; the KYC study must not be merged into the separate account-key harvesting operation.
Sources and citation
Material revision history
- Nov 15, 2026 · Canonical change recorded · new in release · revision 250