The EU has always found a way through. On AI, it’s struggling to see it.
There might be an AI apocalypse. European leaders don’t really want to talk about it.
By SEBASTIAN STARCEVIC and PIETER HAECK
in DUBLIN
Illustration by Natália Delgado/POLITICO
The European Union wasn’t set up to deal with challenges as vast and existential as the possibility of AI wiping out humanity.
For two decades, the EU’s leaders have muddled through crisis after crisis — from Greece coming within days of being pushed out of the eurozone, to Russia’s invasion of Ukraine, which forced millions to seek safety in the bloc. Each time, it has found a way through. It’s what the EU does.
The EU has always found a way through. On AI, it’s struggling to see it.

With AI, though, it’s turning out to be different.
Tech experts, industry and even governments themselves describe paralysis, distraction and defeatism in Europe’s capitals. It’s telling that it took four years after the launch of ChatGPT in November 2022 for EU leaders to bring themselves to discuss artificial intelligence at a summit in Dublin on Friday.
And even then, there was little enthusiasm for the topic.
“There are much more important things,” Latvian Prime Minister Andris Kulbergs said dismissively on his way into the meeting, referring to the pressing, albeit probably not civilization-ending, matter of who should speak for the EU in Russia-Ukraine peace talks.
The authors of a report published two months ago — among them two founders of modern AI and senior executives of OpenAI and Anthropic — would beg to differ with Kulbergs’ assessment. An AI “intelligence explosion” risks “severely eroding checks on power within and between states, companies, and branches of government,” it warned.
Yet even the question of what Europe should actually do about AI remains deeply contentious. Should it try to compete with Washington and Beijing by building its own frontier models, or concentrate on regulating the technology and making sure European businesses and economies can benefit from it? The two goals are not necessarily incompatible, but pursuing them requires money, political commitment and a willingness to move quickly.
European Parliament President Roberta Metsola addressed the continent’s defeatism directly on Friday. “I hate the mantra that Europe has somehow lost the AI race,” she said.
What apocalypse?
If Europe believes it’s on the brink of an AI apocalypse, it is not acting like it.
In Dublin, leaders heard from, among others, Nobel Prize-winning computer scientist Geoffrey Hinton, one of the godfathers of modern AI and a co-author of the intelligence explosion report. Hinton has warned that AI development could outpace humanity’s ability to control it, putting the chances of the technology eventually causing human extinction at between 10% and 20% over the coming decades.
The discussions veered between the existential threat to humanity and the potential economic benefits of the technology, according to an EU official. Leaders also heard from Mistral CEO Arthur Mensch, who has pushed back strongly in recent weeks against the idea that Europe should abandon the race to develop frontier AI models.
The decision to bring in outside speakers was itself an acknowledgement of how little expertise leaders believe they have on the subject. Summing up their relative inexperience on the matter, a senior EU official admitted: “We are not going to pretend that we are experts [on AI].”
Before long, Friday’s conversation moved on to more familiar ground: Ukraine, and who should speak for the bloc in peace talks with Moscow and Kyiv.
The contrast with the United States is striking. In Washington, electoral politics and the growing sums of money pouring in from both pro- and anti-AI groups have forced politicians to develop increasingly substantive positions on the technology.
In Europe, by contrast, the debate has yet to acquire the same political urgency, with the issue too often competing for attention with the bloc’s many other crises and priorities. The leaders will next return to the topic at a summit in three months’ time, European Council President Antonio Costa said. Greek Prime Minister Kyriakos Mitsotakis wants to get all Greek public servants familiar with the basic functions of artificial intelligence — in the next three or four years.
But the consequences of falling behind in AI may be every bit as political as they are economic or technological. The leaders gathered in Dublin, among them French President Emmanuel Macron and German Chancellor Friedrich Merz, have already witnessed the fraying of Europe’s political consensus and the rise of nationalist challengers. Marine Le Pen’s National Rally is positioning itself for power in France, while Alternative for Germany is attracting support on a scale not seen by a far-right force in the country since the 1930s. Economic disillusionment and anger over migration are helping fuel that shift.
AI threatens to add another layer of disruption, with potentially profound consequences for employment, information and democracy.
European AI CEO Uljan Sharka has warned that the bloc is particularly vulnerable to interference in elections, with several key EU countries heading to the polls next year. He has called for a system of monitoring and responding to AI-driven electoral interference, arguing that it should be treated like “terrorism.”
There is also the more immediate question of whether Europe can function without the technology and infrastructure it imports from abroad.
A Microsoft outage on Thursday and Friday affected EU institutions, offering an uncomfortable reminder of the bloc’s dependence on American technology. Asked what would happen if Europe failed to wean itself off that dependence, one Brussels official offered a blunt assessment.
“We’re doomed.”
Giorgio Leali and Nektaria Stamouli contributed to this report.
Source - www.politico.eu ↗