London, 13 September 2026 — EBM Weekend Read — By Brad Adams
In 2020, Palantir Technologies accepted a fee of £1 to help the NHS manage data during the Covid pandemic. Five years later, the company has accumulated more than £670m in current and past UK government contracts spanning the NHS, defence, policing, the nuclear weapons agency AWE and, since March 2026, the Financial Conduct Authority. The journey from a symbolic £1 foothold to a position deep inside some of Britain’s most important public institutions is remarkable, but it is also revealing. Palantir’s strategy has often been described as “land and expand”: get the technology into an organisation, demonstrate that it can solve a difficult problem, and then build a much larger relationship around it. Britain has become perhaps the clearest European example of that approach, and the important question now is not simply whether Palantir makes effective software. It almost certainly does. The more difficult question is what happens when governments become so dependent on a private company’s technology that leaving becomes considerably harder than joining in the first place.
What Palantir Actually Sells
Join The European Business Briefing
New subscribers this quarter are entered into a draw to win a Rolex Submariner. Join 40,000+ founders, investors and executives who read EBM every day.
SubscribeAlex Karp is an important part of that story. He is an unusual figure to be running one of the world’s most valuable software companies, with a law degree from Stanford and a PhD in philosophy from Goethe University Frankfurt, and he has said that he never learned to code. His public persona has been equally unconventional. Karp has embraced his reputation for eccentricity, has acknowledged that colleagues have privately described him in extreme terms and has repeatedly defended Palantir’s military contracts without much interest in making the company more palatable to its critics. He has argued that technology companies should be prepared to support Western governments and national security interests rather than pretending that technology can exist outside politics. His own political journey has been more complicated. Karp was previously a significant Democratic donor and described himself as progressive, while co-founder Peter Thiel became one of the most prominent conservative figures in the technology industry. Karp has subsequently moved closer to the politics of national security and has spoken positively about Donald Trump’s approach to government and immigration enforcement. None of this proves anything about Palantir’s technology, but it does help explain why the company has been willing to occupy a more politically exposed position than many of its Silicon Valley peers.
Britain Became the Test Case
Britain’s relationship with Palantir follows an unusually clear progression. The company entered through healthcare during the Covid crisis, expanded into policing, moved further into defence and has now established a presence in financial regulation. The NHS relationship eventually developed into a £330m contract to run the Federated Data Platform, putting Palantir’s software at the centre of how sensitive health information is managed. In 2025, during Donald Trump’s state visit to Britain, the UK announced a £1.5bn defence partnership with Palantir as part of a wider tech prosperity deal, including plans for the company to establish Britain as its European defence headquarters and create hundreds of jobs. The Ministry of Defence subsequently awarded Palantir a further £240m contract without a competitive tender. There is a perfectly rational argument for this. Britain needs to modernise its armed forces, improve its ability to use data and AI, and work with companies that can actually deliver those capabilities at scale. The problem begins when the immediate benefits of a contract obscure the longer-term consequences of building critical government systems around one supplier.
That tension is particularly obvious in the issue of technical lock-in. A government can theoretically switch software suppliers whenever it wants, but that becomes much harder once employees have been trained on a particular system, databases have been structured around it, workflows have been redesigned and decision-making processes depend upon its architecture. The initial procurement decision may therefore look like an ordinary commercial transaction, while the second and third contracts are made in a completely different environment. The supplier is no longer simply selling software; it has become part of the organisation’s infrastructure. This is why the scrutiny surrounding Palantir’s UK contracts has become more intense. Questions have been raised about the company’s links with former government officials, procurement processes and the possibility of long-term dependency. None of those issues automatically establishes wrongdoing, and it would be wrong to imply otherwise. But collectively they raise a legitimate policy question: at what point does a government stop being a customer with choices and become an organisation that cannot realistically afford to leave?
Europe Is Taking a Different View
The answer varies considerably across Europe. Germany offers an instructive example because individual state police forces have used Palantir’s technology while the federal military has taken a much more cautious position on procurement from US technology companies. Switzerland has also rejected Palantir on national security grounds. France provides the more interesting case because it is attempting to develop a domestic alternative after using Palantir technology within its domestic intelligence infrastructure since the terrorist attacks of 2015. Paris has selected French technology company ChapsVision to develop a replacement, an attempt EBM has examined in detail. The French experience illustrates the central difficulty with the idea of digital sovereignty. It is relatively easy for politicians to say that Europe should control its own critical technology. It is considerably harder to replace a system that is already functioning, has been integrated into government databases and has become familiar to the people who use it. A domestic alternative can take years to develop, while the foreign system is already delivering results. That creates a powerful incentive to stay with the incumbent, even when policymakers have reservations about doing so.
The argument against Palantir therefore goes beyond the familiar criticism of “Big Tech.” It concerns the particular combination of healthcare, defence, intelligence, policing and government data that now surrounds the company. Critics have raised concerns about Palantir’s NHS work, its involvement in military targeting and its work with US immigration authorities, while the company’s relationship with the Israeli military has generated further controversy. There are also legitimate questions about procurement and the movement of personnel between government departments and private contractors. But there is a danger in reducing the debate to a simple argument about whether Palantir is a good company or a bad company. Its critics can point to genuine concerns about privacy, accountability and dependency, while its supporters can point to software that governments actually want to use and capabilities that are difficult to reproduce quickly. The more important issue is what happens when those two realities collide. A government can believe that Palantir offers the best technology available today and still conclude that relying on one American supplier for critical infrastructure creates a strategic vulnerability tomorrow.
That dilemma has become more acute because Europe’s need for technology is growing faster than its ability to produce alternatives. The war in Ukraine has demonstrated the importance of real-time intelligence, data integration and software-enabled military decision-making, while the rapid development of AI has made data infrastructure a strategic concern rather than merely an IT issue. European governments are therefore being asked to make a choice between capability now and sovereignty later. Palantir has a substantial advantage because its systems already exist, have been tested in demanding environments and can be deployed without waiting for a European competitor to mature. The commercial market has recognised the opportunity too. Palantir’s share price has risen dramatically as investors have priced in accelerating government and commercial demand. That success is not difficult to understand. What governments are buying is not simply software; increasingly, they are buying speed, capability and an established technological ecosystem at a moment when building an alternative could take years.
The Bigger Question for Europe
The Palantir debate is ultimately part of a much larger European problem. Governments across the continent talk about technological sovereignty while remaining heavily dependent on American chips, cloud infrastructure and AI systems for the underlying technology stack. Palantir simply makes the issue unusually visible because its technology sits so close to the most sensitive functions of the modern state. A government can rationally buy American technology because it is cheaper, faster or better. But once that technology is embedded in healthcare, defence, intelligence or financial systems, the procurement decision stops being purely commercial. It becomes a strategic decision about who controls the infrastructure through which the state itself operates.
That does not mean European governments should automatically reject Palantir. There is a serious case for using technology that works, particularly when national security is involved and European alternatives are not yet capable of doing the same job. Nor should every contract involving former officials or limited competition be treated as evidence of misconduct. The more interesting question is whether governments are thinking far enough ahead when they sign these deals. A £1 pilot does not create dependence. One defence contract does not create dependence either. But a sequence of contracts across healthcare, policing, defence and finance can gradually create something much more difficult to unwind. By the time policymakers recognise the strategic significance, the technology may already be too deeply embedded for a clean exit to be realistic.
That is why Palantir’s European expansion deserves attention. The company does not need to control governments to acquire considerable influence over how they operate. It only needs to become useful enough, quickly enough and deeply enough embedded that replacing it becomes more painful than continuing with it. Europe has spent years talking about sovereignty, but sovereignty is ultimately tested not by speeches or strategy papers but by whether a government can walk away from a supplier when it wants to. Palantir’s success suggests that Britain, and increasingly other European countries, may be approaching the point where that question becomes much harder to answer. The choice is not necessarily between Palantir and European independence. It is between using the technology while retaining credible alternatives, or allowing today’s useful supplier to become tomorrow’s indispensable infrastructure. That distinction could prove considerably more important than the next contract Palantir wins.
