Lately, the U.S. and the U.K. seem to be moving closer than ever. Donald Trump, who has always pushed his “America First” slogan, acts like a different person when he is in Britain. Surrounded by tech and finance giants, he announced a wave of investments in the U.K., unprecedented in scale: £150 billion in total, or about $200 billion.
Microsoft promised to spend £30 billion over the next four years on hyperscale data centers. Nvidia, about £11 billion, including direct investment in British companies and supplying 120,000 of its new B200 GPUs. Google, another £5 billion to expand its local data center and deepen DeepMind’s cooperation with the British government. Palantir pledged £1.5 billion for AI-powered defense systems. CoreWeave, with Nvidia being its significant investor, also announced £1.5 billion for an AI data center. Even Blackstone, a private equity giant, plans to pour in £90 billion mainly for data centers in the U.K.
On top of this, both countries signed the UK-U.S. Tech Prosperity Deal, a framework for deeper cooperation in AI, quantum, nuclear energy, and other key areas.
The British government says these projects will create more than 5,000 jobs and inject new energy into the economy. That’s not hard to believe. Data center construction is one of the most certain growth opportunities of the next five years, pulling along energy, chips, power grids, servers, even water systems.
But here’s the question: Why is America being so generous? Is it about helping Britain kickstart an “AI infrastructure boom”? Or is it about forging an AI alliance with the U.K.?
I don’t think it’s either of those, at least not at the core. Behind the show of generosity, Trump and the American tech giants are pursuing something else: a new round of global digital expansion. These companies, holding both capital and core technologies, build their presence through large-scale data center investments. What they’re really building is their own AI empire. “Mutual benefit” sounds nice, but in reality, this is America extending its reach.
The Business Opportunity Beneath AI Infrastructure
For companies like Microsoft and Blackstone, it makes perfect business sense. The new AI boom is like a fresh Industrial Revolution. Back then, power belonged to whoever controlled railroads, ships, and electricity networks. Today, the scarce, strategic resource is compute power, and data centers are its core.
By building these facilities in the U.K., American corporations lock themselves in as long-term service providers, earning steady profits from local and global customers.
And data centers are not easy to replicate. They’re expensive to build, difficult to operate, and naturally monopolistic, like power grids, highways, or water utilities. Once a company plants itself, competition becomes almost impossible for latecomers.
Even more, unlike railroads or highways, these AI facilities can’t simply be handed over to anyone. They must be run by the companies that built them. Which means the core of Britain’s digital economy may eventually sit in American hands.
A New Colonial Logic for the AI Era
Tech companies aren’t harmless little rabbits. They’re powerful, resource-hungry actors that can influence policy wherever they land. And AI companies, in particular, crave data and compute. The bigger the scale, the smarter the models, and the stronger their market position. It’s a textbook Matthew Effect, where the strong keep getting stronger. That’s why these giants are so eager to capture larger and more diverse data from around the world to feed their models.
Many countries don’t allow sensitive data to leave their borders, for reasons of sovereignty and privacy. So, the tech giants came up with a workaround: build data centers locally, and deploy AI services directly on-site. The data stays in the country, at least technically.
If you think about it, doesn’t this resemble the logic of old empires? Once upon a time, controlling trade routes and chokepoints, like the Strait of Malacca or the Suez Canal, meant controlling global trade, and by extension, pricing power. Today, it is data centers.
And data centers are more subtle. The data may stay within the country, but the operation of those data centers is in foreign hands, techniques like federated learning make it possible to train and optimize models across borders. The result: Britain’s most strategic resource, its data, quietly flows outward, while everything looks peaceful on the surface.
Handing over data centers to foreign companies comes with with long-term risks. In the short term, Britain gets jobs and investment. But long term? Its AI industry, and maybe its whole digital future, could be shaped by someone else.
Two hundred years ago, agricultural nations couldn’t resist the cannons and ironclads of industrial empires. Today, facing America’s AI expansion, other countries may find themselves equally powerless.
DeepMind, Britain’s only world-class AI company, is already acquired by Google. Europe’s situation isn’t much better: hundreds of millions of people, plenty of data, but no homegrown internet or AI infrastructure giants. The startup Mistral has promise, but so far it’s still small and almost invisible. Maybe Europe is the next stop on America’s expansion path.
In the middle of the AI wave, only those countries and companies with real strategic vision will be able to see through the patterns beneath the surface, grasp the key technologies, build their own computing infrastructure, and hold a position in the coming AI revolution.