Breaking
Smart Devices

EU Commission hit by AI-powered cyberattack

By Owen Fitzgerald 3 min read
EU Commission hit by AI-powered cyberattack - ai cyberattack
The intrusion was first detected on 22 October and a preliminary assessment finished on 26 October.

The EU institution confirmed a cyberattack that exploited a publicly available AI tool, forcing several services offline. The incident highlighted how tools meant for research can be turned against defenders.

Timeline and immediate response

Officials reported that the intrusion was first detected on 22 October. Within four days, a preliminary assessment was finished on 26 October, allowing the body to gauge the initial impact and plan next steps. After the detection, the organization launched an internal investigation to trace the source and scope of the breach.

As a protective measure, the agency temporarily disabled multiple applications that could be leveraged by the attackers. Most of those services have now been reinstated.

Coordination for the investigation is being handled together with CERT-EU and the European Data Protection Supervisor, according to statements from the agency’s communications office.

AI tool behind the breach

A separate security report identified the software as ARTEX, an AI-powered penetration-testing framework. The original repository for this framework was taken down in October after multiple reports of its misuse in hostile operations.

Data impact and ransom claims

Investigators have not found evidence that staff personal data was compromised at this stage. The spokesperson said no data leak has been confirmed, and no unauthorized extraction of files has been detected.

When asked about attribution, the representative declined to comment, stating that the probe is ongoing and that motives remain unclear.

Related Post: Sony-owned ISP data breach hits 2.6 million users

Details from the Commission’s Press Briefing

Commission spokesperson Thomas Regnier confirmed the breach during Thursday’s midday press briefing in Brussels. Politico was the first outlet to publish an exclusive report on the attack earlier that day.

Regnier declined to name the AI tool involved and did not answer a reporter’s request to verify whether it was a Chinese system. He also refused to state whether the attack had concluded, noting that many elements were contained. Regnier said definitive conclusions would wait until the investigation is finished.

When asked about the Commission’s own defensive AI tools, Regnier said security reasons prevented him from detailing the measures or software used. He added he could not speak for other EU bodies, limiting comments to the Commission alone. CERT-EU identified the attackers as using ARTEX, and a journalist queried why the Commission remained silent on a similar incident weeks earlier until Politico broke the story.

Commission’s stance on AI and security

The spokesperson reiterated that while AI can help detect vulnerabilities, it can also be weaponized, and “no one is immune to this.” He referenced the July action plan on cybersecurity and AI as evidence that the Commission is actively monitoring emerging threats.

He explained that keeping defensive measures out of public view is intended to preserve their effectiveness. The steps taken were described as following the usual procedural steps, designed to limit exposure during an active investigation.

When pressed to label the incident’s magnitude, the official avoided a definitive judgment, describing it simply as a campaign.

The investigation is being coordinated with CERT-EU and the European Data Protection Supervisor, ensuring that technical expertise and data-privacy oversight are combined throughout the response effort. The Commission will likely provide further updates once the investigation is complete.

Owen Fitzgerald

Leave a Reply

Your email address will not be published. Required fields are marked *