Q-Cyber Magazine · Incident report

Government · 23 November 2026 · 6 min read · Updated

European Commission confirms AI-assisted cyber campaign that took staff systems offline

The European Commission says an AI-generated cyber campaign, detected on 22 October 2026, led it to take several internal applications offline. Whether staff data was taken is still under investigation.

Lit office windows across the facade of a multi-storey building at dusk

Brussels, 23 November 2026. The European Commission has confirmed that it responded to what its spokesperson called "an AI generated cyber campaign", detected on 22 October 2026, by temporarily taking several internal applications offline. Speaking at the midday press briefing on 19 November, the spokesperson for digital, Thomas Regnier, said most of the applications had since been restored and that the investigation was continuing.

What happened

Asked about a Politico report published that morning, Regnier said the Commission "has recently taken necessary steps in response to a detected cyber campaign. As a protective and preventive measure, we temporarily took down several applications, and most of them have since then been restored." He said the incident was detected on 22 October and confirmed on 26 October, after which the Commission opened an investigation and took the preventive measures he described.

Regnier confirmed that "this was an AI generated cyber campaign" and "a misuse of an AI tool", and said the tool "is publicly available". He declined to describe the scale of the attack while the investigation continues, but said: "It is a campaign, so there was a series of incidents." In a statement to Politico, he said that individual systems were targeted and that those concerned had been informed in line with standard practice.

Which systems were affected

The Commission has not listed the affected applications. According to Politico's Brussels Playbook, which first reported the incident, disruption lasted over ten days. Politico reported that Sysper, the Commission's HR platform for payslips, leave requests and timesheets, and MyPMO, its medical reimbursement system, had been restored earlier that week, that Sysper briefly crashed again on 18 November, and that translation services were also hit. Salaries were paid as usual, according to the officials Politico spoke to, but some staff could not submit timesheets or request leave.

On 19 November, two EU officials told Euronews that Sysper was working again but that the Joint Sickness Insurance Scheme (JSIS) remained unavailable. Euronews noted that the HR platforms hold personal details of officials, including names, pay grades, home addresses and contact details, and that the Commission's internal documents are hosted on a separate system.

What is confirmed, and what is not

Confirmed by the Commission on the record: the detection and confirmation dates; that the campaign was AI-generated and used a publicly available AI tool; that several applications were taken down and most restored; that the EDPS was informed; that the Commission is in close contact with CERT-EU; that the spokesperson was not aware of any ransom demand; and that no evidence has been found of a data breach affecting staff as a whole. In an email reported by Infosecurity Magazine, a Commission spokesperson added that it is "preventively checking other systems".

Reported, not confirmed by the Commission:

  • The tool. Politico reported that an EU official said CERT-EU had detected that "international hackers" used ARTEX, adding: "Some actors misuse such tools and have targeted the Commission." Regnier did not name the tool.
  • The internal response. Politico reported, citing a leaked internal mail, that the Corporate Management Board, the Commission's highest internal crisis-management body, met on the "ongoing cyber incidents", and, citing two officials, that Executive Vice-President Henna Virkkunen briefed the College of Commissioners.

Not yet known: who is behind the campaign, which Regnier declined to discuss ("on the attribution, I will not go there"); whether the campaign is over; and whether the personal data of any individual staff members was accessed. The Commission's wording refers to staff "as a whole".

What CERT-EU has said about ARTEX

On 5 November, CERT-EU, the cybersecurity service for the Union institutions, bodies, offices and agencies, warned that several threat actors around the world had used ARTEX to automate their campaigns. It describes ARTEX as an open-source, LLM-driven autonomous penetration-testing platform that rapidly identifies and exploits misconfigurations as well as traditional software vulnerabilities. According to CERT-EU, the original author made the repository unavailable in October 2026, and the versions now available are independent forks.

CERT-EU said it had "observed similar instances of AI-driven activities against its constituency", which "highly likely involved the use of ARTEX". Its post does not name the Commission. CERT-EU assessed that the tool's public availability "will likely accelerate adoption by threat actors, resulting in more frequent, faster, and more scalable automated attacks with a higher success rate".

The official response

The Commission says it took the applications down as a protective and preventive measure, opened an investigation and is "taking any necessary steps based on the findings as they come". The spokesperson declined to describe the defensive measures or tools being used, for security reasons. Asked why the Commission had not announced the incident earlier, he said measures taken against a cyber campaign are not made public so that they remain effective, that the usual information-sharing processes were followed and that "there was no hiding whatsoever".

What it means

  1. Automation shortens the time between exposure and exploitation. CERT-EU's guidance starts with the basics: know which assets are exposed to the internet and limit that exposure, keep management and control panels off the internet, and patch exposed systems first. It also advises keeping even non-severe findings in the remediation cycle, because AI-enabled tools can chain minor misconfigurations into higher-impact attack paths. An independent cybersecurity audit is one way to find those gaps before an automated tool does.
  2. Privileged access needs watching, not only granting. CERT-EU recommends multi-factor authentication on all critical systems and close monitoring of privileged and administrator accounts. Knowlcom's zero-trust architecture work and the article on zero trust for government networks cover per-request verification and logging.
  3. HR systems hold targeting data. Names, home addresses, pay grades and health-insurance details are exactly what a convincing phishing message needs. Organisations should keep HR data to what is required, control and log who can read it, and prepare staff for targeted messages while an investigation is open. See data security, and the Denmark CPR breach for why knowing someone's details is not proof of identity.
  4. Plan the outage and the message. Taking applications offline was the Commission's protective measure; during the disruption, some staff could not submit timesheets or request leave, Politico reported. Decide in advance which systems can be isolated, what the secure manual fallback is, and when and how staff and regulators are told. The article on incident reporting for critical operators sets out the reporting side, and CERT-EU's advice that security monitoring should also cover AI tools applies to defenders' own AI deployments.

The investigation is continuing. This article will be updated if the Commission or CERT-EU publish further findings.

Sources

  1. European Commission Audiovisual Service, EC Midday press briefing of 19/11/2026: Cyber campaign against the European Commission, Q&A with transcript (19 November 2026)
  2. CERT-EU, From Penetration Testing to Cyber Threat: Beware of ARTEX (5 November 2026)
  3. Politico Brussels Playbook (Gerardo Fortuna), Cyberattack hits the Commission (19 November 2026)
  4. Euronews (Luca Bertuzzi), EU Commission hit by large-scale cyberattack with data leakage unknown (19 November 2026)
  5. Infosecurity Magazine (James Coker), European Commission Confirms AI-Driven Cyber-Attack Took Systems Offline (19 November 2026)
  6. CSO Online (Evan Schuman), European Commission HR hit by cyber attack (19 November 2026)