The Pulling of Mythos Offline: Why AI KYC Will Fail to Stop Cybercriminals

The report warns that U.S. export controls forcing AI labs to implement financial‑grade KYC will be circumvented by established cybercrime ecosystems: infostealer malware (Lumma, Vidar, RedLine) supplies session cookies and credentials, underground markets sell pre‑verified accounts and identity fabrication services, and deepfake/voice tools defeat biometric checks—while centralized KYC data collection itself becomes a lucrative breach target that will further enable identity fraud.