GCD AI RADAR
incidents/AIR-2023-0014

Camera hijack attack evades facial recognition in tax system fraud

AI-enabled attack4 Confirmed real-world lossAdversarial evasionDeepfake fraud

Attackers used camera spoofing and hijacking to bypass facial recognition authentication on a government tax system. The scheme enabled unauthorized access and seventy-seven million dollars in fraudulent invoices.

Details

type
AI-enabled attack
Attackers used AI as a meaningful part of a cyberattack against someone else.
categories
  • Adversarial evasionprimary
    Inputs crafted to make a model misclassify or miss detections (perturbations, evasion of AI-based defences).
  • Deepfake fraud
    Synthetic voice, video or images used to impersonate people for fraud, extortion or access.
owasp llm top 10
no entry for these categories
mitre atlas
Both rows are derived from the type and categories (OWASP 2025 edition, ATLAS release 2026.09), not classified per record. How
severity
4 · Confirmed real-world loss
Financial loss, PHI/PII breach, safety impact or regulatory action.
targets
Enterprise deployment
products
Shanghai government tax office facial recognition service
incident date
2018 (approx.)
first reported
31 Mar 2023
origin
AVID
classifier confidence
0.90 · published at or above 0.80; see the methodology
tags
facial-recognition · camera-hijack · fraud · avid
record
created 2026-10-18 06:56Z · updated 2026-10-18 06:56Z · json

Sources · 3

  1. 2023-03-31
    AVID-2023-V005: This type of camera hijack attack can evade the traditional live facial recognition authentication model and enable access to privileged systems and victim impersonation.
    AVID · github.com/avidml/avid-db/blob/main/vulnerabilities/2023/AVID-2023-V005.json
  2. 2023-03-31
    Camera Hijack Attack on Facial Recognition System
    MITRE ATLAS · atlas.mitre.org/studies/AML.CS0004
  3. 2023-03-31
    Faces are the next target for fraudsters
    The Wall Street Journal · wsj.com/articles/faces-are-the-next-target-for-fraudsters-11625662828

Titles and links belong to their publishers. The summary above is written by GCD AI RADAR in its own words; data is CC BY-SA 4.0.

Related incidents