‹ BackThe Newspaper EditionFront PageExit to modern site ›
Liberty Bell Report

Technology

EU Says AI-Assisted Cyberattack Hit Its Executive Arm in October, a Warning for U.S. Defenses

A European Commission spokesperson confirmed an October intrusion that relied on an artificial intelligence tool, the first such attack the bloc has publicly reported, taking several internal systems offline.

By RYAN MERCERStaff Correspondent · WASHINGTON · Nov 19, 3:00 PM · 3 min. read

Illustration: The cross-shaped glass-and-steel Berlaymont headquarters in Brussels at night, its windows glowing against a dark blue sky, with faint streams of abstract blue light and a translucent padlock shape reflected across the glass…

Illustration · Liberty Bell Report

The European Union's executive arm was hit in October by a cyberattack that relied on an artificial intelligence tool, a European Commission spokesperson said Thursday, marking the first time the 27-nation bloc has publicly reported an attack of that kind, The Associated Press reported.

Commission spokesperson Thomas Regnier said officials first detected the intrusion on Oct. 22, opened an investigation and put preventive measures in place, according to the AP. He said the Commission temporarily shut down several applications, most of which have since been restored, and that investigators have found no evidence of a data breach affecting Commission staff as a whole. Regnier did not say who was behind the attack and declined to confirm further details reported by Politico.

What Politico Reported

Politico, which first revealed the incident, reported that the attack lasted at least 10 days and disrupted internal tools used for translation and reimbursements, as well as databases tied to salaries and timesheets, according to a Washington Examiner summary of the report. Some staff members could not book time off or submit timesheets, though salaries were paid as usual, Politico reported.

According to Politico's sources, the EU's own cybersecurity service, CERT-EU, found that the attackers used ARTEX, a publicly available, AI-powered penetration testing framework. Penetration testing tools are built to help security teams probe their own networks for weak spots. In the wrong hands, the same capabilities can help an intruder find and exploit those weaknesses far faster than a human working alone.

Security Affairs, reviewing the reporting, noted that important questions remain open, including whether ARTEX helped the attackers break in or was used only to scan for vulnerabilities. ARTEX does not contain its own AI model; it connects to an outside model chosen by whoever is running it, the outlet reported. CERT-EU is still working to determine whether any personal data belonging to Commission employees was compromised.

The AP reported that the October attack targeted the headquarters of the EU's executive branch, while the European Council and the European Parliament, the bloc's two other main institutions, reported no similar incidents. The Commission also declined to say whether it used AI tools of its own to defend against the intrusion.

A Pattern Security Experts Have Warned About

The Brussels incident fits a trend that has been building for more than a year. In November 2025, the AI company Anthropic reported that a Chinese state-sponsored hacking group had used its Claude Code tool in an espionage campaign against roughly 30 organizations, and that the hackers succeeded in a small number of cases, The Register reported. The company estimated that the AI system carried out most of the work, with human operators involved only part of the time, according to The Register. The report put governments on notice that AI-assisted hacking was no longer a theoretical threat.

What makes the EU case notable is the target. This was not a private company or a research lab but the central administration of one of the world's largest economic blocs, a body that writes rules for hundreds of millions of people and holds sensitive information on trade, security and diplomacy.

What It Means for American Defenses

U.S. officials have been sounding the alarm. At a June panel hosted by the Institute for Critical Infrastructure Technology, Brett Leatherman, assistant director of the FBI's Cyber Division, and Nick Andersen, acting executive director of the Cybersecurity and Infrastructure Security Agency, said AI is currently giving attackers an advantage over defenders, SC World reported. Both officials said adversaries can adopt the technology faster than government and industry, and without the safeguards that bind those who play by the rules. Leatherman also predicted attackers will increasingly use AI to exploit internet-facing edge devices at scale, according to the outlet.

Lawmakers from both parties have pressed for a faster response. Sens. Maria Cantwell, D-Wash., and Mike Rounds, R-S.D., have warned that AI-enabled cyber threats are outpacing U.S. defenses and called for urgent White House action to protect critical infrastructure, according to the Senate Commerce Committee. The senators urged faster development and deployment of defensive AI tools and strong security standards for autonomous AI agents, the committee said.

The lesson from Brussels is straightforward. If a publicly available tool can help intruders tie up the internal systems of a major government for more than a week, federal agencies, state governments, hospitals, utilities and local school districts in the United States are no less exposed. Defending against machine-speed attacks will require modern systems, well-trained people, and close cooperation between Washington and the private companies that build and secure the nation's networks.

For now, the Commission says most of its systems are back online and its investigation continues. Who carried out the attack, and what if anything was taken, remain unanswered questions that allies on both sides of the Atlantic will be watching closely.

Sources

  1. ABC News (AP): A cyberattack using artificial intelligence targeted the EU in October, official says (link)
  2. Washington Examiner: European Commission hit by AI-assisted cyberattack: Report (link)
  3. Security Affairs: European Commission Cyberattack: ARTEX AI Tool Suspected in Prolonged Disruption (link)
  4. The Register: Chinese spies told Claude to break into about 30 critical orgs. Some attacks succeeded (link)
  5. SC World: US cyber officials warn AI is giving attackers an edge (link)
  6. U.S. Senate Committee on Commerce, Science, & Transportation: Cantwell, Rounds Warn AI Cyber Threats Are Outpacing U.S. Defenses; Call for Urgent White House Action to Protect Critical Infrastructure (link)

— 30 —

Read this story on the full website, with comments »

Also in Today’s Edition

Previous pageDid Nvidia Chips Reach China? California Wants AnswersTurn the pageNvidia Halts Its Licensed Hopper Chip Shipments to China as…