العربية
🚨 BreakingFBI Warns Artex AI Tool Automates Hacks at Machine Speed

AntiFox AI Pentest Tool Used to Steal AWS Production Secrets

Security World Teamاقرأ بالعربية

Arctic Wolf observed attacks using AntiFox, an autonomous AI pentesting platform built on Artex, to steal AWS credentials and move through compromised environments, in some cases reaching production secrets.

Illustration for Cybersecurity News
Illustration by Security World

⚡ Why it matters

Offensive AI tools compress hours of work into minutes, and one AWS key leaked from a config file can open dozens of cloud services to an attacker.

AntiFox is a GitHub project that bills itself as an agentic pentesting platform built on Artex. A multi-agent system runs a full engagement, from recon through exploitation to a written report. Arctic Wolf says Alibaba Cloud hosts serving AntiFox interfaces were tied to confirmed malicious activity.

Attack path

  • Entry: some intrusions began with a compromised internet-facing IIS app, where the attacker dropped an ASPX web shell via w3wp.exe; the exact flaw is unknown. Others started with a valid AWS credential of unknown origin.
  • Recon: PowerShell commands hunted for web.config files and config backups and read database connection strings.
  • Cloud: valid IAM identities were used to enumerate dozens of services, including EC2, S3, Lambda, Secrets Manager, KMS and RDS. Most calls failed with AccessDenied, but some succeeded, including a GetSecretValue call against a production secret.

AWS said its services were not compromised and that attackers used credentials taken from customers' compromised applications.

What to do

  1. Keep AWS keys out of config files. Use IAM roles and Secrets Manager with least privilege.
  2. Enable CloudTrail and alert on bursts of AccessDenied errors and unusual GetSecretValue calls.
  3. Monitor IIS servers for new ASPX files and processes spawned by w3wp.exe.

🔗 Sources

This article is original reporting by the Security World editorial team based on the sources listed above; direct quotes are kept short and marked with quotation marks. Corrections and feedback: info [at] securitywd.net.

Related stories