FBI Warns Hackers Are Weaponizing ARTEX AI Pentesting Framework to Automate Attacks Worldwide

The FBI has warned that multiple threat actors are using ARTEX, an AI-powered penetration testing framework, to automate cyberattacks against organizations worldwide. In a FLASH advisory released Friday, the bureau said ARTEX has been weaponized since at least August 2026 against the financial, government, technology, defense, healthcare, and other sectors, including in the recent ConnectWise supply-chain breach that compromised Microsoft 365 customer environments.
According to the FBI, ARTEX greatly shortens the time needed to chain attack techniques. Some intrusions went from initial targeting to privileged access in under three hours. The framework also lets attackers with little technical skill run complex operations against many targets at once.
What is ARTEX?
ARTEX is an autonomous penetration testing framework written in Go. It uses AI agents to carry out reconnaissance, vulnerability discovery, exploitation, privilege escalation, and data theft with little human involvement.
- No built-in AI model: ARTEX connects to external large language models (LLMs), including Anthropic Claude, OpenAI GPT, DeepSeek, GLM, Grok, and Kimi.
- Goal-driven: Attackers give ARTEX their goals and tasks, and it turns them into commands carried out by specialized AI agents.
- Scale: It can automate intrusions against multiple organizations at the same time.
- Targets overlooked systems: The FBI says it is especially good at finding weaknesses in partner portals, employee applications, legacy services, and third-party integrations.
These automated attacks often generate a lot of suspicious traffic that can trigger alerts. But they move so fast that defenders may have very little time to respond before a system is breached.
Still widely available despite takedown
A developer known as Autumn-27 released ARTEX on GitHub in July 2026 and removed the repository on October 8, after reports that threat actors were abusing it. The FBI says it is still available through:
- Community forks
- Archived copies
- Localized versions
- A commercial edition called ARTEX PRO
As of early November, researchers had found more than 1,600 publicly exposed ARTEX consoles, and the real number of deployments is believed to be much higher.
Known ARTEX-linked attacks
| Campaign | Threat actor | Impact |
|---|---|---|
| ConnectWise Cloud Backup supply-chain attack (October 2026) | Storm-0940 (China-state-aligned, per Microsoft) | Application identities of ConnectWise’s cloud backup service (formerly SkyKick) compromised. Authentication attempted against 2,250+ downstream Microsoft 365 tenants; Huntress found at least 162 compromised tenants |
| South Korean financial institutions | Chinese-speaking threat actor | Combined ARTEX with Claude Code agents against banks including Shinhan, KB Kookmin, and Hana. Personal data of about 67,000 people exposed. CrowdStrike found the attacker’s exposed infrastructure |
| South Korean criminal gang | Seven suspects arrested by South Korean police | Attacks on 90+ organizations, including hospitals, schools, and businesses. About 8.2 million data records stolen |
| Automated espionage campaign | Vapor Panda (Chinese cyberespionage group, per CrowdStrike) | 200+ organizations in 14 countries targeted; at least 9 compromised |
In the ConnectWise attack, Microsoft says Storm-0940 used ARTEX to speed up reconnaissance, password spraying, and abuse of cloud application permissions. Some intrusions went on to change credentials and try to set up persistent access. In the Vapor Panda campaign, AI agents performed reconnaissance, exploited vulnerabilities, deployed webshells, stole credentials, and tried different attack paths when their first attempts failed.
Detection resources
- FBI advisory: The 13-page FLASH includes indicators of compromise (IOCs) from known ARTEX campaigns, including IP addresses, command-and-control infrastructure, and suspicious network activity. The FBI notes that the listed IP addresses may have changed since the attacks.
- MalwareMustDie signature pack: Researcher MalwareMustDie released an open-source ARTEX detection pack on Friday. It has rules and indicators for security monitoring tools, covering suspicious network requests, identity abuse, cloud activity, malicious files, and automated attack patterns, along with a playbook that ranks indicators by confidence level. The researcher warns that infrastructure indicators such as IP addresses can go stale or cause false positives.
Why it matters
ARTEX shows how AI agents are changing the threat landscape. Attacks that once took skilled operators days or weeks can now be automated, run in parallel across hundreds of targets, and adapt in real time. With less skill needed and attacks moving faster, organizations need detection and response that works in minutes, not hours.
FBI recommendations
- Secure internet-facing applications, including partner portals, employee apps, and legacy services.
- Restrict access to cloud administration interfaces.
- Review application permissions and authentication logs, especially for third-party apps with access to your cloud tenant.
- Rate-limit sensitive APIs.
- Monitor for:
- Unusual bursts of automated requests
- Suspicious cloud activity
- Unexpected credential or privilege changes
- Large data transfers
- Report suspicious ARTEX-related activity to the FBI or IC3.
Additional steps for Microsoft 365 environments
- Audit enterprise applications and service principals in Entra ID, and remove unused or overly broad permissions granted to third-party apps.
- Check whether you used ConnectWise/SkyKick Cloud Backup, and review sign-in logs for that application’s identity.
- Detect password spraying: Watch for many failed logins across accounts from the same sources, and enforce phishing-resistant MFA.
Read the full FBI FLASH: https://www.ic3.gov/CSA/2026/261113.pdf



