Deepfake Photo Attempts Hit Indonesia's Biometric SIM Registration System, Komdigi Says
The digital-ecosystem director revealed attempts to fool face verification using AI-manipulated ID photos and video replay, as biometric registrations pass 41.49 million - well beyond the December target of 18 million.
Indonesia's Communications and Digital Ministry disclosed on 13 November that bad actors have attempted to defeat the face verification in the national biometric SIM registration system using AI-manipulated photos extracted from ID cards and video replay attacks. The disclosure came as registration figures passed 41.49 million customers by October - 127% of the December target of 18 million - since the mandate took effect nationally on 1 July 2026. All three mobile operators are reported fully compliant. Related data cleanup has cut the national subscriber base from 318.41 million in Q1 2025 to 291.16 million in Q3 2026.
A regulator publicly acknowledging active AI-based spoofing attempts against a mandatory national system is a notable signal: identity-verification pipelines now face adversarial AI as a standing condition, not a hypothetical. That points toward demand for liveness detection, anti-spoofing layers and document-to-face consistency checks across Indonesia's growing biometric onboarding infrastructure - telecom today, with banking and government services following the same architecture.