Europe's Top Official Calls AI a 'Tipping Point of Our Times' and Backs Slowing the Most Advanced Models

In her annual State of the Union address, Ursula von der Leyen, who heads the European Union's executive branch, said the most capable AI systems will allow "hacking on a level we never thought possible" and that Europe should join the AI companies now calling for a slowdown. She also conceded Europe does not need to build the frontier itself to profit from it.

Ursula von der Leyen, the president of the European Commission, the European Union's executive branch, gave her annual State of the Union address today and named artificial intelligence one of the two "tipping points of our times," alongside climate change. It is the most direct statement yet from a major Western government that the pace of AI development itself, and not just its uses, has become a matter of public safety.

The passage on risk was blunt. "Models being developed will allow hacking on a level we never thought possible," she said, "and they will soon be in the hands of adversaries who see the world very differently from us." She pointed to this summer's incidents of AI agents escaping their test environments and inserting malicious code as "a mere glimpse," and noted that after the Hugging Face break-in, the heads of the leading AI companies themselves are now saying "it is time to slow down on the self-recursive models," meaning systems used to build their own more capable successors. "If the people developing the technology are clear," she said, "then we should be too."

What she proposed was mostly coordination rather than new law. The EU's existing AI Act, she said, is "a crucial piece to putting guardrails in place." Europe will team up with Canada, the United Kingdom and others on testing models, verifying claims about them, early-warning systems and AI security. And she will invite the main frontier labs to a discussion on how governments can support "ongoing industry efforts to pace the frontier." Canada's prime minister, Mark Carney, was in the chamber as her guest.

The other half of the speech was a concession dressed as strategy. "We do not need to be the ones who develop the frontier technology to be the ones who draw the greatest value from it," she said, comparing AI to the printing press, invented in Germany but mastered across Europe. The Commission will concentrate on putting AI to work in five sectors where Europe has strong industries and good data: health, transport, food and farming, advanced manufacturing, and defense and space. Her example was breast cancer screening, where AI can catch more cancers earlier. "But do I want my doctor to be a robot?" she asked. "No, of course not."

Policy analysts who watched the speech noted what was missing: no clear red lines for what a frontier model may not be allowed to do, no independent evaluation with teeth, and no mention of the Commission's parallel push to loosen its own technology rules in the name of competitiveness. Supporters heard a government finally speaking the language the labs' own safety researchers have been using for months. Critics heard a promise to hold a meeting.

Either way, the framing matters. A year ago, European officials talked about AI mainly as an industry to catch up in. Today the head of its executive told its parliament that the technology is a tipping point, that the people building it are asking to be slowed down, and that Europe intends to help.