Europe Cannot Afford to Miss the Intelligence Era

This week, London made one of the most consequential technology decisions of the AI era.

Sadiq Khan, the Mayor of London, blocked the Metropolitan Police’s proposed £50 million contract with Palantir. The Mayor’s Office for Policing and Crime cited “serious concerns” with the procurement process — not on political grounds, not on civil liberties grounds, but on a more revealing basis: the Met could not evidence it had considered any other supplier.

One supplier. For a £50 million intelligence contract covering criminal investigations across one of the world’s largest police forces.

That single procurement finding is a window into something far larger than a contract dispute.

It is an early signal that governments, enterprises, and institutions are beginning to recognise that the intelligence layer itself has become a sovereignty issue — and that, for many institutions, the consolidation has already quietly happened.

How We Got Here

For years, Europe gradually outsourced the foundational layers of the digital economy.

First came search. Then social. Then cloud infrastructure.

Today, American hyperscalers reportedly control nearly 70% of Europe’s cloud infrastructure market, while European providers have declined to roughly 15% market share.

Now, that dependency is extending beyond storage and compute into the intelligence layer itself.

Palantir alone is reportedly embedded across more than ten UK government departments — spanning healthcare, defence, policing, and local government — through contracts worth hundreds of millions of pounds.

The Met’s response to the Mayor’s veto was telling. The force called the decision “disappointing,” stating it “needed to use the best technology available.” It was an admission, perhaps unintentional, that alternatives had not seriously been evaluated — not because they do not exist, but because the dependency had already become the default.

Notably, neither side has ruled out resuming the arrangement. The Mayor’s Office indicated it remains open to a revised procurement process. The contract may yet proceed. But the episode surfaced something important: when an institution responsible for the security of ten million people cannot articulate why it evaluated only one intelligence supplier, the question of who governs the intelligence layer has already been answered — by default, not by design.

That pattern is not unique to the Metropolitan Police. It is an early warning sign of a much larger transition: the operational intelligence layer of Western institutions is consolidating into externally governed systems — quietly, contract by contract, department by department.

The Sovereignty Shift

As AI systems become increasingly operational — coordinating workflows, analysing sensitive information, supporting decision-making, and eventually acting autonomously across critical infrastructure — the question is no longer simply what these systems can do.

The question is: who controls the environment they operate inside?

Unlike previous software eras, AI systems are not passive tools. They are becoming operational actors inside governments, financial institutions, telecommunications providers, healthcare systems, defence organisations, and enterprise infrastructure itself. That changes the nature of trust entirely.

Europe spent the last decade debating data sovereignty. The next decade will be defined by intelligence sovereignty.

The cloud era centralised compute. The intelligence era risks centralising cognition itself.

Yet much of today’s AI ecosystem still depends on externally governed infrastructure layers — centralised hyperscalers, opaque orchestration frameworks, and intelligence environments enterprises ultimately do not own or control. That may work for consumer applications. It does not work for the future of sovereign enterprise infrastructure.

The Economic Argument

This is not only a sovereignty question. It is an economic growth question.

The consequences of missing the intelligence era will not be confined to security briefings or national security reviews. They will show up in GDP figures, capital markets, and the long-term trajectory of European prosperity.

Consider the London Stock Exchange. Once among the world’s top five exchanges by market capitalisation, the LSE has now slipped outside the global top twenty. Companies that would previously have listed in London are choosing New York, where the technology sector — built on the very infrastructure layers Europe ceded — dominates global valuations and attracts global capital. The exchange that once reflected the breadth of British economic power now reflects, in part, the downstream cost of two decades of technological dependency.

This is not coincidence. It is causation.

When the foundational layers of the digital economy — search, social, cloud — are built and governed elsewhere, the economic value generated by those layers accrues elsewhere too. Talent migrates. Capital follows. Companies scale within ecosystems that control their own infrastructure and list on exchanges that reflect those ecosystems’ valuations.

The intelligence era presents precisely the same choice, at even greater stakes.

Artificial intelligence is not simply another software category. It is rapidly becoming the primary mechanism through which enterprises generate productivity, compress decision cycles, and create durable competitive advantage. The organisations — and by extension the economies — that govern their own intelligence infrastructure will compound that advantage over time.

Those that remain dependent on externally governed intelligence environments will face the same fate as European cloud providers: declining market share, diminishing leverage, and an increasingly structural disadvantage in global capital markets.

The economic growth potential of the intelligence era is enormous. But that growth will not be distributed evenly. It will concentrate in the ecosystems that control the foundational layers — precisely as the economic value of the internet concentrated in the ecosystems that controlled search, social, and cloud.

Europe has an opportunity to change that trajectory. But only if it builds.