Ledger / Nov 2026
Attackers use autonomous AI (Cairn) against semiconductor, defense and civil-society sectors
Confirmed
Unit 42, which investigated, states the attackers paired Cairn with GLM and DeepSeek.
- Model
- GLM, DeepSeekZhipu AI · open-weight; DeepSeek · open-weight
- Agent harness
- Cairnopen-source harness
- Attributed country
- Not attributedthe source names no country
- Sectors hit
- manufacturing, defenseSemiconductor, defense and civil-society organizations (25 operations)
Attackers weaponized the open-source Cairn agent framework with Zhipu GLM and DeepSeek models across 25 operations against semiconductor, defense and civil-society targets.
“The attackers weaponized Cairn by pairing it with Chinese AI models Zhipu GLM and DeepSeek.”Palo Alto Networks Unit 42
Sources
- Attackers use autonomous AI (Cairn) against semiconductor, defense and civil-society sectorsPalo Alto Networks Unit 42 · Nov 2026
- Date
- Nov 2026
- Actor
- Unnamed operators using Cairn (unknown), as attributed by Palo Alto Networks Unit 42
- Targets
- Semiconductor, defense and civil-society organizations (25 operations)
- What the AI did
- Agentic intrusion, Recon, Exploit dev
- Model as named
- GLM, DeepSeek
- Model basis
- A researcher names it
- Autonomy
- Ran on its own
- Outcome
- 25 operations against semiconductor, defense and civil-society organizations.