top of page

The AI Hacking Tool ARTEX Hit Europe. The US Too. 5 Ways to Protect Your Business

2 hours ago
7 min read

The video uses a digital avatar of Erick Grau, CEO of Chibitek.

This is not just a European story, and it is not a story about some new kind of attack. It is a story about speed, and about a door most businesses forget they have.

ARTEX is an AI hacking tool. The FBI describes it as "an open-source, large language model (LLM)-driven agentic platform that threat actors have weaponized." In plain terms, an attacker types a goal in plain language and a team of AI agents does the scanning and breaking in. It was published as a penetration testing tool, its developer pulled the original repository in October, and the FBI says it is still widely available through mirrors.

If you followed our earlier post on AI gateway keys, ARTEX will sound familiar. This post covers what has happened since, and what to do about it.

What Happened in Europe

On November 19, the European Commission confirmed an AI-driven attack on its HR systems. At the Commission's midday briefing, a spokesperson said, "we can confirm that this was an AI generated cyber campaign," calling it "a misuse of an AI tool." The Commission itself did not name the tool on the record. Euronews and CSO Online both report that the tool was ARTEX.

Here is what is reported:

  • The attack was first detected on October 22, according to Euronews and CSO Online.

  • As a precaution, the Commission took several internal applications offline. CSO Online and The Register both report more than 10 days of disruption. One of the affected apps was Sysper, which staff use for salary information, timesheets and vacation requests.

  • Most of those apps have since been restored, according to the Commission spokesperson quoted by CSO Online.

  • The Commission said, "we have not detected any evidence of a data breach affecting Commission staff as a whole." It also said it was not aware of any ransomware request.

The Commission declined to say who was behind the attack, so we won't guess either.

What the FBI Says About the US

In the US, the FBI's November 13 FLASH alert is the clearest picture of scope. It says that "since at least August 2026, a variety of malicious cyber actors have used ARTEX to conduct campaigns targeting organizations globally," across sectors that include financial services, government, technology, defense, healthcare, manufacturing, telecommunications and legal.

Two numbers from that alert stand out:

  • Some intrusions went from first targeting to privileged access "in under three hours."

  • As of early November, researchers found "over 1,600 publicly exposed ARTEX console instances."

Three hours is shorter than a long lunch. If your plan is that someone will notice an alert on Monday, that plan does not fit this attack.

So Where Does "The US Too" Come From?

We want to be exact here, because this is where headlines get sloppy. The FBI describes ARTEX campaigns against organizations "globally." The most specific US detail comes from CrowdStrike.

CrowdStrike reports that a China-nexus group it calls Vapor Panda used ARTEX against more than 200 organizations in at least 14 countries, beginning October 17. In its words, "A separate engagement targeted 19 U.S.-based organizations in the pharmaceutical, semiconductor, defense, nuclear, and financial sectors. The adversary also compromised telecommunications providers in the U.S., Belgium, and the Philippines."

CrowdStrike: Vapor Panda. 200+ organizations, 14+ countries. 19 US organizations targeted in pharma, semiconductors, defense, nuclear and finance. US telecoms compromised.

So: 19 US organizations were targeted, and US telecom providers were compromised. That is what "the US too" means in this post, and no more.

The Two Doors

The most useful lesson for business owners comes from a different case: the ConnectWise Cloud Backup compromise in October. Microsoft attributes it to a China-state-aligned group it tracks as Storm-0940, which it says used ARTEX, and BleepingComputer and Huntress report the same attribution. CrowdStrike says its Vapor Panda activity is separate from what industry reports as Storm-0940, so treat these as two different stories.

Think of your Microsoft 365 tenant as having two doors.

The people's door. This is your staff signing in. Attackers tried this door too: Huntress reports password spray attempts against staff accounts in the days before the app sign-ins. But the unauthorized access Huntress went on to confirm came through a different door.

The app's door. This is every app and vendor you have connected to Microsoft 365. Huntress found that a credential belonging to the backup app, an API secret, had been compromised. With it, the attackers signed in as the app itself, using permissions customers had already granted it. The app signs in with its own credential, not a staff member's password. The FBI says the compromise led to attempted sign-ins to more than 2,250 downstream tenants, and Huntress confirmed unauthorized sign-ins in 162 tenants.

WaterISAC put it well: "ARTEX does not introduce new techniques," and "What changes is the speed and scale." One compromised vendor can open a path into hundreds of customer environments.

Most businesses have spent years locking the people's door. Far fewer have looked hard at the app's door.

5 Ways to Protect Your Business

The first three steps line up with the FBI's and Microsoft's published recommendations. The last two are Chibitek guidance.

1. Count Your App Doors

List every app and vendor connected to your Microsoft 365 tenant. Backup, CRM, email tools, that pilot from two years ago. Remove what you don't use. The FBI recommends applying least privilege to multi-tenant app permissions and consent grants, and the FBI and Microsoft both advise reviewing every tenant where an affected app is or was installed, even if the integration is no longer in use.

2. Make Keys Expire

Every app secret should have a named owner and an end date. When a vendor contract ends, revoke that vendor's access the same day. Microsoft and the FBI both say to revoke unnecessary permissions and rotate credentials when something looks wrong. We would rather you not wait for something to look wrong.

3. Watch App Sign-Ins, Not Just People

Most alerting watches people. The FBI and Microsoft both recommend reviewing service principal sign-ins, which are the app's sign-ins, and the FBI also recommends monitoring "credential additions, privilege assignments, and anomalous app-only activity." At minimum, set an alert for when a new key is added to an app. Huntress saw exactly that move in one tenant, where an abused app added new credentials to other apps.

4. Keep the People's Door Locked

Password spraying is still part of the playbook, as Huntress saw in this campaign. Keep MFA on every account and keep account lockout turned on. Watching the app's door is no reason to loosen this one.

5. Run a Friday Test

Plant a fake key, often called a canary token, that alerts when someone uses it. Trip it yourself on a Friday night and time how long it takes for a real person to respond. If the answer is Monday morning, you found that out on your terms instead of an attacker's.

The Bottom Line

ARTEX doesn't bring new tricks. It brings speed. The European Commission story shows the disruption can last for days, the FBI shows the reach is global, and CrowdStrike shows US organizations are in the mix. The ConnectWise case shows where many businesses are exposed: not the people's door, but the app's door.

Key Points

  • The FBI says attackers have used ARTEX against organizations globally since at least August 2026, with some reaching privileged access in under three hours.

  • The European Commission confirmed an AI-driven attack on its HR systems and found no evidence of a breach affecting staff as a whole. Euronews and CSO Online report the tool was ARTEX.

  • CrowdStrike reports one group targeted 19 US organizations and compromised US telecom providers.

  • In the ConnectWise case, attackers used the app's own credential, not anyone's password.

  • Count your app doors, make keys expire, watch app sign-ins, keep MFA and lockout on, and test your response time.

Frequently Asked Questions

Is ARTEX a new type of attack? No. WaterISAC notes it does not introduce new techniques. What it changes is how fast and how widely attackers can work.

We use MFA. Are we covered? MFA helps protect the people's door. App credentials, like the API secret in the ConnectWise case, sign in without a user and without an MFA prompt. You need to watch both.

Who attacked the European Commission? The Commission has not said, and we won't speculate.

Do I need new tools to do this? Not to start. Steps one through four are mostly about reviewing what is already in your Microsoft 365 tenant and turning on the right alerts.

Where can I read more about stolen app secrets? Our post on stolen Entra app secrets walks through how to check yours.

Let's Count Your App Doors

At Chibitek, we help businesses keep technology effortless, with no geek-speak. If you're not sure how many apps are connected to your Microsoft 365 tenant, or who would notice if one of them started acting strange on a Friday night, contact us at hello@chibitek.com and we can walk through it together.

Lock the people's door. Then count the app doors.

We make IT effortless so you can create, disrupt, and grow!

Follow Chibitek on YouTube at @chibitekmsp, and follow Chibitek on LinkedIn.

Sources

All pages were reopened on Nov. 20, 2026, and every quote above was found word for word.

  1. FBI, FLASH-20261113-001, "Threat Overview of ARTEX Autonomous Penetration Testing Framework," November 13, 2026 (PDF)

  2. CrowdStrike, "Vapor Panda Uses ARTEX to Scale Global Campaign"

  3. Microsoft Threat Intelligence, Storm-0940 and the ConnectWise supply chain compromise, October 21, 2026

  4. Huntress, "ConnectWise (SkyKick) Vendor Credentials Abused in M365 Attacks," October 16, 2026

  5. European Commission, midday briefing, November 19, 2026 (video and transcript)

  6. Euronews, EU Commission cyberattack report, November 19, 2026

  7. CSO Online, "European Commission HR hit by cyber attack," November 19, 2026

  8. The Register, European Commission AI-assisted cyberattack, November 19, 2026

  9. WaterISAC, FBI warning on ARTEX

  10. BleepingComputer, FBI warning on ARTEX, November 15, 2026

  11. CyberScoop, Vapor Panda ARTEX campaign

  12. The Record, ConnectWise Cloud Backup supply chain attack

  13. Chibitek, "The AI Hacking Tool ARTEX Hit Europe. The US Too." (YouTube Short)

Comments


bottom of page