Hunt: ARTEX indicators from the FBI FLASH and Amazon's MadPot, and the behaviors that give agentic intrusions away
- FBI, ARTEX host (South Korea campaign):
38.244.50[.]120; proxies101.53.80[.]20,205.214.59[.]31,124.155.252[.]63,154.201.79[.]246,23.248.249[.]90,23.158.220[.]98,103.248.148[.]84,203.160.133[.]172,209.209.85[.]38; LLM reseller domainxcai[.]pro. - FBI, ConnectWise campaign:
144.172.64[.]60,.99,.128(sign-ins and Graph API);187.15.88[.]186,187.15.90[.]217,187.15.89[.]186(Graph via compromised apps, Oct 15-16). - FBI, Storm-3358 (active as of Nov 6): WISP C2
199.19.95[.]66-68; ARTEX servers91.92.240[.]238,85.11.168[.]45,178.16.54[.]129,31.77.229[.]35; C285.11.167[.]52. - Amazon MadPot (Oct 14 to Nov 2): AI-credential theft
104.28.225[.]223,31.22.91[.]156,150.136.38[.]110,128.242.115[.]131,216.176.239[.]231; KYC-deepfake campaign212.107.28[.]51,103.142.140[.]158,104.28.251[.]98,104.28.251[.]99,104.28.219[.]97,125.103.213[.]41. - Behaviors (FBI): sub-second Azure capability checks and test blob write/delete, ARM deployment validation dry-runs and AWS DryRun permission probes; lookups to
api.db-ip[.]com; command output sent to rotatingwebhook.siteURLs; unexpected egress to LLM APIs or self-hosted gateways (CLIProxyAPI, New API); WISP, vshell, AdaptixC2 and BLOODALCHEMY; new 100 GB volumes before exfiltration; bursts of near-duplicate requests from sub-agents overwriting each other. - AI apps (Amazon): agents fingerprint One API/New API/Veloera/Sub2API/LiteLLM gateways and Dify, FastGPT, Langflow (CVE-2025-3248), RAGFlow (CVE-2026-28797), Open WebUI, LobeChat and Ollama; check for unclaimed installs, default LiteLLM master keys and One API
isCliRequestheader abuse.
Sources: FBI FLASH-20261113-001: Threat Overview of ARTEX Autonomous Penetration Testing Framework (TLP:CLEAR, Nov 13, 2026) · AWS Security Blog: Amazon Threat Intelligence observes threat actors running cybercrime campaigns with ARTEX AI agents (Nov 13, 2026)