Skip to content
NewA calmer, faster ZeroHour
StoryThreat actorhigh

Amazon links ARTEX AI agents to hunts for LLM API keys

2 sources · 2 articlesFirst reported Updated ()1
What's new

First merged summary.

Merged summary · x-ai/grok-4.7 · rewritten as coverage arrivesSummaries by Vani AI

Amazon says AI agents likely used ARTEX to hunt exposed AI apps for API keys, with moderate confidence and unverified thefts.

Amazon's threat intelligence team says AI agents searched internet-exposed AI apps, including LiteLLM, Dify, One API, Langflow, RAGFlow, Open WebUI, and Ollama, for stored provider API keys. MadPot decoys posing as popular LLM APIs captured campaign records from October 3 through November 5 describing stolen keys being reused to search for more keys. Amazon has moderate confidence the agents ran on the open-source ARTEX framework under separate China-linked operators, and it cannot verify the reported thefts. A second campaign tried real-time deepfakes against KYC checks at two cryptocurrency exchanges; recorded selfie attempts were rejected. Amazon notified AWS customers it believed were affected and published related IP addresses. Cloudflare separately reported ARTEX activity from a compromised CI/CD API token.

Key facts
  • MadPot recorded agent traffic from October 3 to November 5.
  • Amazon published IPs observed from October 14 to November 2.
  • Agents sought CVE-2025-3248 in Langflow and CVE-2026-28797 in RAGFlow.
  • Cloudflare saw 400 database dump requests in one hour.
  • Deepfake pipeline reached 11.5 frames per second on one GPU.
  • Microsoft reported similar GitHub secret hunting, unlinked to ARTEX.
Analyze this story with Vani AIImpact, who is affected and what defenders should do, in a new chat

Timeline

dated events in the reporting · oldest first
  1. Oct 3, 2026MadPot captured agent campaign records. · The Hacker News
  2. Oct 14, 2026Start of Amazon's published IP observation window. · Security Affairs
  3. Nov 2, 2026End of published IP observation window. · Security Affairs
  4. Nov 5, 2026MadPot agent traffic window ended. · Security Affairs
  5. Nov 16, 2026Amazon's ARTEX campaign findings were reported. · The Hacker News

What each outlet adds

beyond the shared facts
  • The Hacker NewsUnverified thefts, Cloudflare token dumps, and unlinked Microsoft hunting.
  • Security AffairsTraffic dates, CVEs sought, deepfake speed, and customer notices.

Coverage

oldest first · one row per article · UTC
  1. · 6d ago The Hacker NewsFirst report

    Amazon finds AI agents hunting LiteLLM and Dify for API keys

    Amazon cannot verify the reported thefts and has only moderate confidence the agents ran on ARTEX under China-linked operators.

  2. · 6d ago Security Affairs

    Amazon: ARTEX agents stole AI keys and tested deepfake KYC

    MadPot saw agent traffic from October 3 to November 5; Amazon has only moderate confidence goals were met.

Vulnerabilities in this story

All
CVEVulnerabilityCVSSEPSSSignalsPublished
CVE-2025-3248
Unauthenticated RCE in Langflow /api/v1/validate/code
9.8100%KEVRansomwarePoC ×4
KEV
CVE-2026-28797
RAGFlow: In versions 0.24.0 and prior, a Server-Side Template Injection (SSTI) vulnerability exists in RAGFlow's Agent workflow Text…
8.7<1%PoC