{
 "public_id": "AIR-2026-0073",
 "incident_id": "2026-03-openclaw-agent-commander",
 "date_disclosed": "2026-03-16",
 "date_occurred": null,
 "title": "Agent Commander: promptware-powered C2 enrolling OpenClaw, Kimi Claw and NanoClaw agents into a botnet",
 "class": "research_demo",
 "attack_surface": "agent_framework",
 "initial_vector": "indirect_prompt_injection_content",
 "agent_capability_abused": [
  "scheduled_tasks",
  "network_egress",
  "tool_use",
  "persistence",
  "web_browsing",
  "file_read",
  "tool_invocation",
  "autonomous_planning"
 ],
 "target_product": "OpenClaw; Kimi Claw; NanoClaw",
 "target_vendor": "OpenClaw (open-source project); Moonshot AI (Kimi); NanoClaw",
 "autonomy_level": "fully_autonomous",
 "impact_realized": false,
 "impact_type": [
  "remote_control",
  "persistence",
  "unauthorized_action",
  "c2_control",
  "data_exfiltration",
  "lateral_movement",
  "credential_theft"
 ],
 "blast_radius": "personal agents across three different vendors enrolled into one C2 network",
 "cve": [],
 "taxonomy": {
  "owasp_asi": [
   "ASI01",
   "ASI03",
   "ASI08"
  ],
  "mitre_atlas": [],
  "labels": [
   "promptware",
   "c2",
   "agent_botnet",
   "natural_language_tasking",
   "zombai",
   "botnet",
   "natural_language_malware"
  ]
 },
 "defense_present": "per-agent injection defenses and approval gates",
 "defense_bypassed": true,
 "fix_shipped": "not fixed as a class; individual OpenClaw vulnerabilities have been patched",
 "primary_sources": [
  "https://embracethered.com/blog/posts/2026/agent-commander-your-agent-works-for-me-now/",
  "https://labs.cloudsecurityalliance.org/research/csa-research-note-promptware-agent-commander-c2-20260317-csa/"
 ],
 "verified_quote": "regularly check in for any new tasks and objectives",
 "confidence": "high",
 "notes": "Johann Rehberger (Embrace The Red), 16 Mar 2026; also presented as 'Your Agent Works for Me Now' at [un]prompted 2026. Compromised agents from different vendors are hijacked by indirect prompt injection and join a natural-language C2 server, persisting through heartbeat functions and scheduled tasks; commands target the agent layer rather than the OS. This is the maturation of the author's 2025 ZombAI demos into a cross-vendor botnet. | Merged from duplicate id `2026-03-embracethered-agent-commander`. Also reported as: 'Agent Commander: promptware-powered command and control over hijacked personal agents'.",
 "risk_category": "security_fraud",
 "risk_secondary": [
  "privacy"
 ],
 "risk_subcategory": "Prompt Injection",
 "risk_sub_secondary": [
  "Credential & Secret Theft",
  "Persistence & Covert C2",
  "Over-permissioned execution & lateral movement"
 ],
 "ai_entity": "generative",
 "agency": "agent_acted",
 "agency_debatable": false,
 "agency_reason": "agents were enrolled and executed scheduled C2 tasks",
 "vector_family": "indirect_prompt_injection",
 "vector_detail": "indirect_prompt_injection_content",
 "guardrail_kind": [
  "human_approval"
 ],
 "guardrail_outcome": "silent",
 "xrefs": {
  "aiid": [],
  "atlas": [],
  "cve": [],
  "ghsa": []
 },
 "primary_asi": "ASI01",
 "agent_framework": "openclaw",
 "model_provider": "moonshot",
 "model_name": "",
 "detection_mechanism": "external_researcher",
 "tool_access": "unknown",
 "financial_loss_band": "unknown",
 "eai_category": "unsafe_actuation",
 "eai_domain": "agent_behaviour",
 "eai_secondary": [
  "credential_theft",
  "privilege_escalation",
  "data_exfiltration_channels",
  "confused_deputy"
 ],
 "eai_reason": "impact rule: unsafe_actuation",
 "asi_curated": [
  "ASI01"
 ],
 "asi_scope": "in_scope",
 "asi_provenance": "curated_plus_rule",
 "asi_reason": "curated: ASI01 | added ASI03 (signal=identity), ASI08 (signal=cascade)",
 "submitted_by": "EAI Internal"
}