Skip to content
NewA calmer, faster ZeroHour
StoryThreat actorhigh

FBI says ARTEX AI pentest tool is automating global intrusions

9 sources · 12 articlesFirst reported Updated ()1
What's new

November 15–16: BleepingComputer added seven arrests and about 8.2 million records; TechCrunch added Barracuda’s credential reset; The Register said AWS rejected every observed deepfake.

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

An FBI FLASH says actors have used open-source ARTEX since at least August 2026, sometimes reaching privileged access in under three hours.

The FBI’s November 13 FLASH alert says cybercriminals and nation-state actors are abusing ARTEX, a Go penetration-testing framework that drives external language models from reconnaissance through data theft. Reported use dates to at least August 2026 across finance, government, defense, healthcare and other sectors. Microsoft links Storm-0940 to stolen ConnectWise Cloud Backup credentials and sign-in attempts against more than 2,250 Microsoft 365 tenants; Huntress reported at least 162 related compromises. The GitHub project was removed on October 8, but forks and commercial ARTEX PRO remain. Infosecurity Magazine and Cisco Talos date publication to July 2026; The Record says August. Several outlets cite more than 1,600 exposed consoles by early November; The Register says Censys counted 1,465 on November 4. Most outlets cite data on about 67,000 people in South Korea; TechCrunch says at least 68,000.

Key facts
  • FLASH-20261113-001 says some intrusions reached privileged access in under three hours.
  • ARTEX has no own model and calls Claude, GPT, DeepSeek, GLM, Grok and Kimi.
  • BleepingComputer says South Korean police arrested seven suspects over more than 90 organizations.
  • Authorities cited about 8.2 million stolen records, separate from the roughly 67,000-person bank case.
  • Barracuda said an unrevoked third-party credential was used and it reset customer credentials.
  • Amazon’s MadPot recorded key-theft and deepfake KYC activity from October 3 to November 5.
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. Jul 1, 2026ARTEX published on GitHub by Autumn-27, per several outlets · Infosecurity Magazine
  2. Aug 1, 2026FBI says malicious use observed since at least August 2026 · BleepingComputer
  3. Oct 3, 2026Amazon honeypots begin recording ARTEX-linked campaigns · The Register
  4. Oct 8, 2026Developer removes the GitHub repository after reported abuse · GBHackers
  5. Nov 4, 2026Censys counts 1,465 exposed ARTEX consoles · The Register
  6. Nov 13, 2026FBI publishes FLASH-20261113-001 on ARTEX abuse · The Record

What each outlet adds

beyond the shared facts
  • BleepingComputerSeven arrests, 8.2 million records, and Huntress’s 162 compromises
  • TechCrunchBarracuda credential reset and a 68,000-person South Korea figure
  • The RegisterCensys count of 1,465 and rejected Amazon deepfake KYC attempts
  • Cisco TalosMore than 2,500 ARTEX-associated infrastructure IP addresses
  • The RecordAugust release date and developer Li Puhua’s October 8 denial

Coverage

oldest first · one row per article · UTC
  1. · 3d ago The RecordFirst report

    FBI says Artex AI tool automates intrusions against critical sectors

    Some intrusions reached privileged access in under three hours; Censys saw over 1,600 exposed consoles.

  2. · 3d ago GBHackers

    FBI warns attackers are using ARTEX for fast multi-sector intrusions

    FLASH-20261113-001 says some intrusions reached privileged access in under three hours, with Chinese models common in deployments.

  3. · 3d ago Cyber Security News

    FBI says attackers use ARTEX agents to reach privileged access in hours

    A November 13 FLASH alert says some intrusions combined stolen credentials, exposed apps, and cloud permissions.

  4. · 1d ago BleepingComputer

    FBI warns ARTEX AI framework is driving global intrusions

    Forks and a commercial ARTEX PRO edition remain available after the GitHub repo was pulled on October 8.

  5. · 1d ago DataBreaches.net

    FBI warns ARTEX AI pentest tool is used in global intrusions

    The FBI says multiple actors have used open-source ARTEX since at least August 2026, including in a ConnectWise supply-chain breach of Microsoft 365 environments.

  6. · 1d ago GBHackers

    ARTEX AI agents automate enterprise attacks with Nmap, PsExec and WinRM

    Tests found agents retry after failures, while Microsoft saw attempts against more than 2,250 Microsoft 365 tenants.

  7. · 23h ago Cyber Security News

    Storm-0940 uses ARTEX to reach privileged access in under 3 hours

    Microsoft saw attempts against more than 2,250 Microsoft 365 tenants via stolen ConnectWise Cloud Backup credentials.

  8. · 23h ago Infosecurity Magazine

    FBI warns ARTEX AI pentest tool used in global intrusions

    Campaigns since at least August 2026 include a South Korea leak of about 67,000 people and access to more than 2,250 tenants.

  9. · 21h ago Malwarebytes Labs

    FBI warns criminals use ARTEX AI agents to speed intrusions

    Some observed attacks reached privileged access in under three hours, with abuse reported since at least August.

  10. · 19h ago Cyber Security News

    Talos says hackers automate attacks with ARTEX, PentAGI, and HexStrike AI

    A November 13 Talos report ties ARTEX to Storm-0940 attacks on over 2,250 Microsoft 365 tenants.

  11. · 17h ago TechCrunch · Security

    FBI says Artex AI tool breached networks in under three hours

    Reported victims include ConnectWise customers, South Korean finance firms, and Barracuda.

  12. · 15h ago The Register · Security

    FBI and AWS say ARTEX agents automate attacks on critical sectors

    Amazon’s honeypots saw credential-theft and deepfake KYC campaigns from October 3 to November 5; forks remain after the repo was pulled.