
ChatGPT Drew the Fake Badge, Spies Brought the Malware: A Deepfake Military ID Phishing Campaign
Ourtake
ChatGPT declined to copy a real military ID, then happily produced a 'mock-up' of one. Refusals only go as far as the framing.
What happened
Genians Security Center detected a spear-phishing attack by a state-sponsored threat group that impersonated a South Korean defense institution. The threat actor used ChatGPT to generate a deepfake military government employee ID card, which was sent as a decoy with a malicious shortcut file. The attack deployed AutoIt and batch-script malware to establish persistence and enable internal data theft.
Sources
- genians.co.kr · AI-Driven Deepfake-Based Military ID Forgery APT Campaign (opens in a new tab)
- fortune.com · North Korean hackers used ChatGPT to help forge deepfake ID | Fortune (opens in a new tab)
- uk.pcmag.com · This North Korean Phishing Attack Used ChatGPT's Image Generation (opens in a new tab)
- techradar.com · Kimsuky uses AI to create fake military IDs, targeting South Korea's defense institutions (opens in a new tab)