Safety

EU confirms AI-assisted
cyberattack, with data
exposure unresolved

The Commission’s disclosure puts AI-assisted attacks in concrete terms. For small B2B firms, the practical lesson is about explaining customer-data safeguards—not claiming immunity.

The European Commission confirmed on November 19 that an AI-assisted cyberattack detected October 22 prompted it to temporarily take several applications offline. The incident’s scale and possible data exposure remain under investigation.

The disruption lasted more than 10 days, affecting payroll, leave requests, medical reimbursements and translation services, according to The Record and The Register. It makes AI-assisted attacks a concrete operational concern, although the role attributed to a specific hacking tool remains distinct from what the Commission publicly confirmed.

European Commission spokesperson Thomas Regnier said at a press briefing that most applications had been restored, Reuters reported. He said there was no evidence of a data breach affecting Commission staff as a whole, declined to identify a suspected attacker and said he was unaware of a ransom demand.

Services return, data questions remain

Euronews identified Sysper, which handles salary information, leave requests and timesheets, among the affected platforms. Two EU officials told the outlet that Sysper was working again on November 19, but the Joint Sickness Insurance Scheme remained unavailable. Salaries were paid on time, The Register reported.

The personnel and health-insurance systems hold names, pay grades, home addresses and contact details. Euronews reported that DG DIGIT, the Commission’s IT department, was investigating whether personal information had been compromised; internal documents were hosted on a separate system.

A Commission spokesperson told The Record that affected individuals had been notified and officials were “preventively checking other systems” while maintaining contact with CERT-EU and the European Data Protection Supervisor.

According to The Register, the Commission convened an emergency meeting of its Corporate Management Board earlier in November. Technology commissioner Henna Virkkunen briefed fellow commissioners on Wednesday.

“There was no hiding whatsoever,” Regnier told journalists, The Record reported. He said publicizing defensive measures could undermine their effectiveness.

What’s known about ARTEX

The Commission publicly described the incident as an “AI-generated cybercampaign.” Regnier didn’t name the tool. The Record cited Politico’s reporting that CERT-EU had identified ARTEX, an open-source AI hacking platform.

Released in July, the ARTEX framework uses commercial AI models to automate finding and exploiting vulnerabilities, according to The Record. The Register reported that CERT-EU warned on November 5 about suspected ARTEX-assisted attacks against organizations under its protection.

ARTEX’s developer removed the project’s GitHub repository in October, though forks remained available. Separately, The Register reported that an FBI warning issued the previous week said researchers had found more than 1,600 ARTEX consoles exposed to the internet. In some attacks described in that warning, intruders obtained privileged access less than three hours after identifying a target.

Those warning details don’t establish what happened inside the Commission.

For owners of small B2B firms, the practical inference is about preparation, not a claim that prospects’ buying behavior has changed. Identifying where customer data sits, reviewing safeguards around relevant systems and vendors, and clarifying incident-response responsibilities may help owners answer security questions during sales and onboarding.

The incident doesn’t establish that customers are changing purchasing decisions. Nor can an explanation of safeguards guarantee immunity. The immediate unanswered question is whether the Commission’s disrupted systems also exposed personal information.

Sources