European Commission confirms AI-assisted cyberattack as some tools remain offline
Officials detected the incident on October 22. The investigation remains open, with no evidence detected of a data breach affecting Commission staff as a whole.
Loading page…
Officials detected the incident on October 22. The investigation remains open, with no evidence detected of a data breach affecting Commission staff as a whole.
Listen to this story
The Commission’s response has shifted from precautionary shutdowns toward recovery and wider security checks: most affected applications are back, but some remain offline as investigators examine the October 22 incident. Officials confirm attackers used AI to find vulnerabilities, but have not identified the attackers, named the software, or established a motive. The Record’s report that CERT-EU linked the activity to ARTEX remains unconfirmed by the Commission, leaving the incident’s scope and tooling unresolved.
The Commission took applications offline as a precaution; that does not establish that attackers had disabled those systems.
Politico reported at least 10 days of disruption to translation and reimbursement tools and salary and timesheet databases, details Regnier declined to confirm.
Officials said they had found no evidence of a breach affecting Commission staff as a whole and were unaware of any ransom demand.
Some European Union digital tools remain offline after an October cyberattack involving AI, the European Commission confirmed on November 19. Spokesperson Thomas Regnier said authorities detected the incident on October 22 and took applications down as a precaution. Most have been restored, but the investigation remains open and the attackers have not been publicly identified.
The shutdowns were a defensive response, rather than a list of systems officials confirmed the attackers had disabled. Regnier did not identify the attacker or explain the operation in detail, according to the Associated Press.
Politico, which first reported the incident, described a wider operational impact. It said the attack lasted at least 10 days and affected translation and reimbursement tools, along with databases used for salaries and timesheets. Regnier declined to confirm those details.
CERT-EU oversees digital security across the bloc’s institutions, which together employ about 40,000 people in Brussels alone. The Commission said it was keeping in touch with that agency as it responded. Its spokesperson also told Recorded Future News that affected individuals had been notified, without publicly identifying them.
Regnier explicitly confirmed that AI tools were used and said such tools can be misused to detect vulnerabilities. The Record described the software involved as publicly available. But officials did not name it or explain the attackers’ motivation. AP described this as the first time the EU had reported a cyberattack using an AI tool.
In its November 19 article, The Record relayed Politico’s account that CERT-EU had identified attackers using ARTEX, an open-source AI hacking platform.
ARTEX uses commercial AI models to automate the discovery and exploitation of security weaknesses, according to The Record. The outlet said the FBI had warned the previous week that criminals and state-sponsored groups were using it against government agencies and critical sectors worldwide. Some intrusions progressed from initial targeting to privileged access in under three hours, the warning said.
we have not detected any evidence of a data breach affecting commission staff as a whole.
Thomas Regnier, European Commission spokesperson, quoted by the Associated Press
Regnier’s statement about staff data was qualified: investigators had found no evidence of a breach affecting staff as a whole. He also said officials were unaware of any ransom demand, according to The Record.
Asked why the incident had not been disclosed earlier, Regnier said publicizing defensive measures could undermine their effectiveness. He rejected the suggestion that the Commission had concealed the incident, telling journalists there had been no hiding.
The response now extends beyond restoring applications. The Commission said it was preventively checking other systems and staying in contact with the European Data Protection Supervisor.
Loading discussion...
Join the conversation
Explain where you would draw the line between operational secrecy and public accountability.
Be the first to share a perspective or an experience.
Reader comments
Newest comments first. Replies stay oldest first.