According to a document obtained by Politico, the European Commission’s IT department wants to use AI to fight back against the most dangerous cyberattacks.
DG DIGIT, the Commission’s IT department, wrote the paper, though it is undated, and it explains how the EU’s executive branch should get ready to defend against AI-driven threats.
According to Politico, the main goal is speed, as the Commission’s cyber team wrote that just spotting an attack is not enough if staff cannot stop it quickly.
DG DIGIT plans to use automatic defences in certain high-risk situations. The details of these cases were redacted in the version shared with Politico.
The paper says the plan will not allow machines to act without oversight. There will be strict controls, and people will remain in charge.
Moreover, the document also outlines the threats it expects and says that advanced AI models will make attacks faster, larger, and harder to detect.
It warns about attacks that could happen outside office hours, spread quickly through systems, or cause major data theft before staff can respond.
According to Politico, the cyber division has requested internal approval to move forward. It is unclear whether the plans are already being used.
The Commission did not reply to Politico’s request for comment.
Governments have been racing to deal with AI’s growing ability to find and exploit security flaws.
In July, OpenAI said its own models had breached Hugging Face after breaking out of a secure test, and incidents like that have led to calls for more caution around the most powerful AI models.
In September, Commission chief Ursula von der Leyen used her State of the Union speech to back an AI slowdown. She said the EU would bring AI labs together to talk about how to pace the frontier.
Get the TNW newsletter
Get the most important tech news in your inbox each week.