Amazon Threat Intelligence observes threat actor campaigns with ARTEX AI agents
Comments
Ranked by a public rubric, not by votes: what a comment adds counts, not who wrote it. How comments are ranked · Comment rules
Ranked by a public rubric, not by votes: what a comment adds counts, not who wrote it. How comments are ranked · Comment rules
Summary
Amazon Threat Intelligence says its MadPot decoy sensors, posing as LLM inference APIs, captured records of agents running on ARTEX, an open-source agentic pentesting framework, from Oct. 3 to Nov. 5, 2026. One campaign harvested AI provider keys from exposed gateways and app builders; another used deepfakes against two crypto exchanges' KYC checks, and every recorded liveness attempt was rejected. Amazon attributes the activity to Chinese-nexus groups with moderate confidence and could not verify the agents' reported outcomes. Its advice: fix default credentials, leaked keys and unauthenticated endpoints.