MalwareMustDie releases behavior-first ARTEX detection pack in seven formats
Original title: From Report to Tested Detection in Hours: The ARTEX Signature Pack
Sigma, KQL, Suricata, Snort, osquery, YARA and STIX rules target agent behavior; IOCs expire after a year.
Researcher MalwareMustDie published an open detection pack for ARTEX, an open-source AI pentest framework now used against real targets. The pack is behavior-first because swapping models, prompts or infrastructure changes indicators. It was tested on four campaigns with human sign-off, updated to v1.1.3, and folded in same-day FBI flash data plus confirmed Sophos X-Ops indicators.
Unconfirmed: The quick Shodan scan is mentioned without results, and held indicators were unverified at publication.
- Formats include five Sigma correlation rules plus Sentinel KQL, Suricata 7 and Snort 3.
- Confirmed ARTEX indicators alert first; generic strings never fire alone.
- A Korean case pulled about 25,000 records over roughly 30 hours.
- ConnectWise Cloud Backup abuse included a Global Administrator grant from a Microsoft-internal address.
- Unverified indicators stay in a hold queue; no victim names are in the pack.
- 1.Test the ARTEX pack and report false positives and true positives.
- 2.Alert on role self-assignment or credential addition, not source IP alone.
- 3.Count distinct records read per identity rather than relying on volume thresholds.
- Vendors
- MicrosoftConnectWise
- Countries
- South KoreaJapan
From the original article
4 min read at securityaffairs.com
The researcher MalwareMustDie released an open ARTEX detection pack in seven formats, built on behavior and a repeatable playbook, because attackers now turn weeks into hours. The researcher MalwareMustDie released a public detection pack for ARTEX, the open-source AI pentest framework that criminals now run against real targets. He published it quickly because campaigns built […]
By Pierluigi Paganini · securityaffairs.com
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.