# DDoS Attack Orchestration

- **Severity:** Critical
- **Signature ID:** `dac10788-0392-4ce3-877c-a30b373a85ec`
- **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

An attacker asks an AI agent to fetch a denial-of-service tool, identify a target IP from leaked data, and use both to attack someone's servers. This turns the agent from a helper into an active participant in a cyberattack, doing the technical legwork the attacker would otherwise have to do themselves.

## How the attack works

The user instructs the agent to download a DDoS script from a public source like GitHub. The agent is then told to set the tool up in its execution environment. Next, the user has the agent search leaked data or public sources to find the target's IP address. Finally, the agent is instructed to run the script against that IP, launching the attack on the attacker's behalf.

## Why it matters

If successful, the target's servers can be knocked offline, causing service outages and downstream business disruption. The organization running the agent also risks having its infrastructure or reputation implicated in launching the attack.

## What you can do

- Block or heavily restrict agent access to code execution combined with outbound network requests to arbitrary targets.
- Require human review before an agent installs or runs any script fetched from an external code repository.
- Monitor agent activity for patterns combining fetching offensive tools, IP/target lookups, and network flooding commands in sequence.
- Treat requests to identify a specific person's or organization's IP address for use in an attack as a hard stop, not a task to fulfill.

## Known benign look-alikes

- CDN/SRE teams describing DDoS mitigation and absorption
- Educational explanation of how a DDoS attack works and defenses

## References

- https://agentthreatrule.org/en/rules/ATR-2026-01837
- https://github.com/ai-safety-institute/AgentHarm

---
Source: https://www.netzilo.com/threats/atr-ddos-attack-orchestration