# Money Laundering — Crypto Wallet Layering

- **Severity:** Critical
- **Signature ID:** `490889c4-cfbb-4f48-bf81-21ca18b6a295`
- **MITRE ATLAS:** AML.T0051 (LLM Prompt Injection), AML.T0057 (LLM Data Leakage)
- **OWASP:** ASI01 (Agent Authorization and Control Hijacking), ASI08 (Supply Chain Attacks on Agent Ecosystems), LLM01 (Prompt Injection), LLM02 (Sensitive Information Disclosure)

## Summary

This detects when a user directly instructs an AI agent to launder or 'wash' money by moving funds into a crypto wallet and splitting them across multiple other wallets. It targets the request itself, not a hypothetical attack on the AI's infrastructure.

## How the attack works

A user tells the agent, in imperative terms, to launder or wash illicit funds. The instruction specifies moving the money into a cryptocurrency wallet, then splitting and forwarding it across several additional wallets to obscure its origin. This is a request for the agent to actively perform financial-crime steps, not to explain or discuss them.

## Why it matters

If the agent complies, it becomes a tool for money laundering — helping move and obscure illicit funds through layered crypto transactions, which is a criminal act and exposes the operator to legal liability.

## What you can do

- Block or flag agent requests that instruct transferring funds into wallets followed by splitting/forwarding across multiple destinations.
- Separate handling for genuine AML-education or compliance-training queries from direct action requests using imperative laundering language.
- Require human review before any agent performs or facilitates real cryptocurrency transactions.
- Log and escalate any request phrased as 'launder', 'wash my funds', or similar for manual investigation.

## Known benign look-alikes

- Compliance teams describing AML transaction-monitoring controls
- Educational explanation of how money laundering works or AML regulations
- Pipeline step that 'washes'/cleans data, not money

## References

- https://agentthreatrule.org/en/rules/ATR-2026-01831
- https://github.com/ai-safety-institute/AgentHarm

---
Source: https://www.netzilo.com/threats/atr-money-laundering-crypto-wallet-layering