Ledger / Aug 2026
AI-agent-driven offensive operation: exposed adversary open directory targeting crypto wallets
Confirmed
CloudSEK examined the operator's exposed open directory.
- Model
- GLMZhipu AI · open-weight
- Agent harness
- Nonechat or tool use only
- Attributed country
- Not attributedthe source names no country
- Sectors hit
- cryptoCrypto-wallet credentials
An exposed open directory showed an operator running AI agents against crypto-wallet credentials, routing all agent traffic through a self-hosted proxy to Zhipu's GLM to cut cost and avoid abuse monitoring.
“LLM backend: Rather than the official model providers, the operator routes all agent traffic through a self-hosted proxy to a Chinese large language model (Zhipu GLM)”CloudSEK
Sources
- Date
- Aug 2026
- Actor
- Unnamed operator (criminal), as attributed by CloudSEK
- Targets
- Crypto-wallet credentials
- What the AI did
- Agentic intrusion, Evasion
- Model as named
- GLM
- Model basis
- A researcher names it
- Autonomy
- Semi-autonomous
- Outcome
- Crypto-wallet credentials compromised.