Subscribe
0%
Illustration for: European Commission confirms an AI-driven cyberattack knocked staff systems offline. What we know, and what's still unclear

European Commission confirms an AI-driven cyberattack knocked staff systems offline. What we know, and what's still unclear

The Commission says attackers misused a publicly available AI tool, reportedly the ARTEX hacking framework, in a campaign it detected on October 22 that disrupted HR and translation services for more than ten days. It has found no evidence of a broad staff data breach but won't discuss attribution, scale or whether the attack is over.

Listen to this article0:00 / 5:12

The European Commission confirmed on Thursday, November 19, that it was hit by what its spokesperson called an AI-generated cyber campaign carried out with a publicly available AI tool. The confirmation came at the Commission's midday press briefing in Brussels, hours after Politico first reported the incident. According to Help Net Security, spokesperson Thomas Regnier said the attack was detected on October 22 and confirmed after an initial investigation on October 26.

Why it matters: it's rare for a major government body to say publicly that attackers used an AI hacking tool against it. It follows weeks of warnings about ARTEX, an open-source AI framework that automates finding and exploiting weaknesses. Politico reported that the EU institutions' cybersecurity service, CERT-EU, identified ARTEX in this attack. The Commission itself did not name the tool.

What the Commission confirmed

From Regnier's briefing, as reported by The Record and Help Net Security:

  • The Commission called it "an AI generated cyber campaign" using "an AI tool that is publicly available."
  • It temporarily took down several applications as a protective measure. Most have since been restored, Help Net Security reports.
  • So far it has found no evidence of a data breach affecting Commission staff as a whole. The Record says it notified affected individuals and is checking other systems as a precaution.
  • Regnier said he was not aware of any ransom demand.
  • The Commission is in contact with CERT-EU and the European Data Protection Supervisor.

What it would not say: who did it, what the motive was, how large the attack was, or whether it has ended. Regnier declined to discuss attribution while the investigation continues. Asked whether the attack was over, he said many elements had been contained, per Help Net Security. He also argued that describing the defensive measures publicly would make them less effective.

What was disrupted

The details come from Politico's reporting, as relayed by The Record and Cyber Security News. The Commission didn't say which services were targeted.

System Reported impact
Sysper (payslips, working hours, leave) Down for over ten days in total; restored earlier this week, then briefly down again on November 18
MyPMO (medical reimbursements) Disrupted
Translation services Disrupted on November 17

The sources disagree on pay. The Record lists payroll among the disrupted systems. Cyber Security News, citing Politico, says salaries largely arrived as usual even though some staff couldn't enter salary details, file timesheets or request leave. No source gives a number of affected employees. According to Cyber Security News, technology commissioner Henna Virkkunen briefed other commissioners on November 18.

The Record describes ARTEX as an open-source AI hacking platform released in July that uses commercial AI models to automate vulnerability discovery and exploitation. It reports that the FBI warned last week that criminal and state-backed groups are using it. According to the FBI, some intrusions went from first targeting to privileged access in under three hours.

Help Net Security notes that CERT-EU published its own ARTEX warning on November 5, calling it an AI-powered penetration-testing framework. CERT-EU said ARTEX's original repository was taken down in October after reports of misuse, but forks are still available. It warned that public availability will likely speed up adoption by attackers.

One open question matters a lot. Cyber Security News says it hasn't been established whether ARTEX helped the attackers get in or was only used to scan for weaknesses afterwards. Saying "AI was used" is very different from saying "AI broke in."

This is also the Commission's third known security incident this year. The Record mentions breaches involving Ivanti software in January and its cloud infrastructure in March. Cyber Security News describes the March incident as tied to the Trivy supply-chain attack and affecting the Europa web platform. Officials have not linked those incidents to this one.

What to take from it

  • For EU staff: the Commission says it has notified affected individuals. If you weren't contacted, there is no official sign that your data was exposed. Still, expect phishing that uses the outage as a lure, especially messages about payslips, leave or medical claims.
  • For security teams: ARTEX campaigns studied so far have mostly exploited ordinary weaknesses rather than exotic zero-days. Our Check Point ARTEX breakdown found hardcoded credentials, unauthenticated APIs and exposed keys. The FBI's three-hour figure means the gap between first probe and serious access is now very short. Check internet-facing apps and admin interfaces with that in mind, and read CERT-EU's November 5 guidance.

What to watch: whether CERT-EU or the Commission publishes technical indicators, whether the European Data Protection Supervisor opens a formal case, and whether investigators confirm ARTEX was used for initial access.

Sources

  1. The Record — European Commission confirms cyberattack, says hackers used AI tool therecord.media
  2. Help Net Security — European Commission confirms AI-generated cyberattack disrupted its systems helpnetsecurity.com
  3. Cyber Security News — Hackers Target European Commission with ARTEX AI Tool in Major Cyberattack cybersecuritynews.com

#European Commission#CERT-EU#ARTEX#agentic AI#cyberattack#EU