Suspected North Korean group weaponized Claude Code settings to deploy a backdoor
ccerrato147 · x · 2026-11-18
Per Google, a suspected North Korean group created fake recruiter profiles with AI-generated photos, tricked crypto company staff into running a "fix" script, then edited Claude Code settings in their code so the assistant downloaded a backdoor on every launch. Also recap: China-based groups used AI for target research and exploit weaponization, and Iran's APT42 uses AI including Gemini to write lures.
Related event: Google Exposes AI-Driven Fake Account Farms and Fake Recruiter Attacks(2 posts)→
More from coding & agent
- OpenAI says start daily coding with GPT-6.2 Sol: near-flagship scores at ~1/5 the cost — OpenAIDevs · 2026-11-18
- GPT-6.2 Astra tops Agents' Last Exam: +5.2 points at under one-third the per-task cost — OpenAIDevs · 2026-11-18
- OpenAI details GPT-6.2: Astra is the new flagship with 20% lower API prices than GPT-6 Astra — OpenAIDevs · 2026-11-18
- GPT-6.2 Astra and Sol now live in Codex and the API — OpenAIDevs · 2026-11-18
- LangSmith adds three controls for sensitive trace data access — LangChain · 2026-11-18
- Tejas Kumar's Daiku agent harness filled out his Shark Tank application—and got him on the show — TejasKumar_ · 2026-11-18