It’s 2031 and the US and China are about to tear Europe into items.
The US ploughed huge sums into datacentres and the EU didn’t. China constructed robots and Europe didn’t. American corporations “restructured” their workflows round AI and fired individuals, whereas EU employees went on lengthy lunch breaks and handed over administrative duties to the AI mannequin Claude.
Now the chickens are coming house to roost. Europe’s financial system is a shambles as a result of it doesn’t have its personal AI. Populism is surging, the euro is wobbling, cyber-attacks are shredding EU companies. Brexit appeared like a good suggestion. It seems to be like the top of the European Union.
That, at the least, is the imaginative and prescient of a speculative thought experiment, called Europe 2031, penned by Brussels-based thinktankers and printed fortuitously at some point earlier than the Trump administration decided to block “foreign nationals” from utilizing a much-hyped AI mannequin constructed by Anthropic, known as Fable.
Within the heady week of G7 talks that adopted, the situation has gone viral – feeding a feverish dialogue of the urgency for EU tech sovereignty. It has been learn by members of the European parliament and, say its authors, was introduced up in observe 1.5 discussions between British and German officers earlier this week.
Its authors say they really feel “vindicated”, by the eye it has obtained and by the truth that certainly one of their predictions – that the US would prohibit international entry to superior AI fashions – seems to have briefly come true. They hope the situation will spur Europe in direction of a dramatic course-correction on AI.
The piece is a part of a burgeoning style of fictional AI doomsday situations, created by obscure figures, which have gained stunning traction amongst policymakers over the previous yr. In 2025 there was AI 2027, a thought experiment which culminates in a superintelligent AI killing all of humanity to make method for extra datacentres; in February, one other speculative situation imagined AI upending the US economy. (The primary was learn by US vice-president JD Vance, the second contributed to a inventory market wobble.)
One complication of all this could be that their thought experiment is at instances primarily based on present developments in AI whose end result is unsure or unsure.
Maximilian Negele contributed to Europe 2031, he says, due to the “unimaginable translation barrier” between Brussels and San Francisco, the place AI is being developed. Previously at US thinktank Rand, he left his job this yr to concentrate on the mission.
“As any individual who travels to San Francisco fairly a bit and talks to individuals there, what is occurring in Europe simply appeared like a slow-moving automobile crash to me,” he says.
The situation unfolds from the attitude of a fictional bright-eyed Brussels staffer, Caroline Dubois, who has a German buddy, Christian Vogt, with a startup in San Francisco. On a go to, she’s impressed by America’s “70 or 80-hour” working weeks and discomfited by the conviction amongst tech bros that every thing is about to vary.
Again in Europe, she works to evangelise her well-meaning bosses concerning the impending AI future – however fails to persuade. There’s an excessive amount of scepticism, and most of the people assume AI is a bubble.
Issues go from there. The People spend enormous sums on a large AI constructing programme – the situation highlights a real-life $100bn (£75bn) deal between OpenAI and Nvidia, the $300bn settlement between OpenAI and Oracle, and “bulldozers” breaking earth in Texas for an AI datacentre. Europeans, in the meantime, put ahead a tepid funding bundle and ignore advisers’ pleas for “a full regulatory carte blanche for datacentre suppliers”.
In a matter of years, America monopolises 70% of the world’s “compute” – the semiconductor chips that fill the datacentres that energy AI fashions. Europe’s financial system is in the meantime gasping for air, principally as a result of its corporations haven’t adopted AI.
As AI-powered cyber-attacks shred European companies and unemployment surges, EU officers scramble to parlay their one final bargaining chip – the Dutch lithography agency ASML, which is important to the manufacturing of AI semiconductors – into concessions from Beijing or Washington. Nevertheless it’s too late. The US deploys highly effective “frontier AI” spy ware and learns the deepest fears of EU officers and in addition which ones are having affairs.
Curtains drop. Christian and Caroline exeunt stage left for a drink. Catastrophe impends.
Sceptical readers would possibly level out that various the eye-popping sums and large tasks that the authors name-check in describing the US’s AI ascent have already fallen aside.
The $100bn settlement between OpenAI and Nvidia, the largest AI deal of final yr, evaporated in February. The $300bn between OpenAI and Oracle seems doubtful, particularly as current stories point out the maker of ChatGPT is still billions of dollars underwater because it burns cash on datacentre infrastructure.
The bulldozers on the bottom in Texas will not be bulldozing very a lot any extra, as OpenAI pulled out of the flagship AI mission to which that second within the situation appears to refer.
The authors are sanguine about these issues. All through the piece, they pre-empt potential objections – reminiscent of AI being overhyped – by suggesting that the hapless European officers have these worries, too, they usually find yourself tragically unsuitable.
“I wouldn’t rule out that there’s some exuberance and that one or two AI corporations would possibly go bankrupt,” says Negele. “However what we wished to get throughout is a common really feel for a model of what we expect will occur.”
He and his co-author, Alex Petropolous, agree that there could possibly be some bumps within the street – together with mounting resistance to datacentres within the US. “I imply, individuals hate AI basically. Lots of people do. Folks hate datacentres. They destroy the panorama. They assist massive tech. It’s a really, very unpopular coverage.”
The authors of Europe 2031 assume that the answer to that is datacentres. Europe must construct extra, quicker, ideally in AI zones the place issues reminiscent of energy and planning will be streamlined and deregulated.
“I feel our view is that the whole datacentre provide is kind of an inelastic provide. So there’ll solely be a restricted variety of datacentres constructed on the planet constructed yearly, and the query is, what number of of these would you like constructed within the US? What number of of these would you like inbuilt Europe?” says Petropolous.
It’s additional value noting that the principle organisation behind the Europe 2031 situation, Arq Basis, primarily based in Brussels, describes itself as “neither an advocacy NGO nor a venture-backed startup” and doesn’t disclose who funds it.
Brussels politicians who learn it, although, could take away a less complicated message: the situation has crystallised a dialog concerning the want for Europe to have technological sovereignty.
“This situation, Europe 2031, I consider that a number of the components they talked about can occur,” says Nicolás Casares, a member of the European parliament from Spain. “However I feel they’re growing – a bit – the alarms so as to name our consideration.”
The US reducing off Europe’s entry to Fable, he says, implies that the EU must ask itself more durable questions on who’s constructing its AI infrastructure and who will profit from it.
“What’s the added worth of getting OpenAI or Anthropic datacentres in Europe?” he says.
“We’re shopping for a story that we want lots of datacentres to not lose the race for AI. However that is loopy … we’re paving the way in which for infrastructure that they’ll use and typically not permit us the potential for utilizing it.”